Transformer Architectures

Follow this topic Ask about this topic

Momentum

148 papers in the last four weeks, up 164% on the four weeks before. 1.5% of all new papers.

Jul 13Week of Sep 28

Latest papers 1,120

May 9, 2026stat.ML

Learning Theory of Transformers: Local-to-Global Approximation via Softmax Partition of Unity

This paper investigates the learning theory of Transformer networks for regression tasks on the compact Euclidean domain

[0,1]^d

and

d

-dimensional compact Riemannian manifolds. We propose a novel constructive approximation framework for Transformers that builds local approximations of the target function and aggregates them into a global approximation via softmax partition of unity. This approach leverages the attention mechanism to achieve spatial localization through affine transformations of the input. The softmax activation plays a crucial role in aggregating local approximations to a global output. From an approximation perspective, we prove that a dense Transformer equipped with only two encoder blocks and standard single-hidden-layer point-wise feed-forward networks can achieve a uniform

\varepsilon

-approximation error for

α

-Hölder continuous functions with

α\in (0,1]

using

\mathcal{O}(\varepsilon^{-d/α})

total parameters. Building upon this approximation guarantee, we establish a near minimax-optimal generalization error bound of order

\mathcal{O}\big(n^{-\frac{2α}{2α+d}} \log n\big)

for the empirical risk minimizer, where

n

is the training data size. The Transformer architecture studied in this paper is dense, shallow and wide, and employs softmax activation and sinusoidal positional encodings, closely reflecting practical implementations.

Zhongjie Shi, Wenjing Liao

School of Mathematics, Georgia Institute of Technology, Atlanta, GA 30332, United States

Transformer Architectures Spatio-Temporal Transformers Feed-Forward Partitioning

May 9, 2026cs.LG

Causal Dimensionality of Transformer Representations: Measurement, Scaling, and Layer Structure

Sparse autoencoders (SAEs) decompose transformer residual streams into interpretable feature dictionaries, yet the relationship between SAE width and causal influence on model output has not been systematically characterised. We introduce causal dimensionality kappa(L, M, T), defined as the effective rank of the expected Jacobian outer product at layer L, and show it can be estimated via the SAE width sweep paired with attribution patching. Across seven SAE widths from 16,384 to 1,048,576 features on Gemma-2-2B layer 12, representational capacity grows 15.6x while causal capacity grows only 4.35x: a robust separation we term the representational-causal wedge. A saturating fit yields kappa-hat approximately 1,990 with kappa-hat / d_model = 0.86 and participation-ratio lower bound kappa_PR approximately 280. Crucially, kappa is invariant to model scaling: Gemma-2-9B and Gemma-2-2B yield identical N_causal = 328 at the same SAE width despite a 3.46x parameter increase (the count is forced to 2% of SAE width by calibration; the substantive empirical claim is shape invariance of the AtP score distribution under matched seq=512 conditions). Across eight network depths kappa is constant while the absolute attribution threshold drops 20x from layer 1 to layer 23. Five controls (architecture invariance, threshold robustness, geometric privilege, synthetic ground-truth recovery, and a four-cell encoder/decoder ablation) pin down what kappa measures and what it does not. Our findings establish kappa as a measurable, model-intrinsic property of transformer layers: sub-linearly recoverable by SAE width, invariant to model scaling, and structured across network depth.

Nilesh Sarkar, Dawar Jyoti Deka

Erdős AI Lab

Sparse Autoencoder Features Transformer Architectures Representational Capacity Scaling

May 8, 2026cs.LG

Scaling Limits of Long-Context Transformers

We study the long-context limit of softmax self-attention with a fixed query and a random context of

n

i.i.d. keys on the sphere, viewing the inverse temperature

β_n

as the scaling parameter that decides whether attention degenerates into uniform averaging or collapses onto the single closest key. We show that the critical scale at which selectivity emerges is determined by the local exponent of the distance-to-query distribution near zero rather than by global features of the context, and scales like

β_n^\ast \asymp n^{2/(d-1)}

for uniform keys on

\mathbb{S}^{d-1}

. Furthermore, we characterize the limiting laws of the ordered attention weights and of the attention output across all regimes of

β_n

: a subcritical regime in which the output reduces to a local average around

q

with explicit deterministic bias and Gaussian fluctuations; a critical regime in which a finite collection of nearest keys retains macroscopic mass without single-key collapse; and a supercritical regime in which all mass concentrates on the closest key. Of notable interest is the subcritical case with identity value matrix where the attention map approximately implements a backward heat equation.

Giuseppe Bruno, Shi Chen, Zhengjiang Lin +2

Linear Attention Efficient Long-Context Inference Gumbel-Softmax Relaxation Long-Context Reasoning Transformer Architectures Selectivity+3

May 8, 2026cs.LG

Transformers Can Implement Preconditioned Richardson Iteration for In-Context Gaussian Kernel Regression

Mechanistic accounts of in-context learning (ICL) have identified iterative algorithms for linear regression and related linear prediction tasks, often using linear or ReLU attention variants. For nonlinear ICL, prior work has related softmax and kernelized attention to functional-gradient-type dynamics, but it remains unclear whether a standard transformer with softmax attention can implement a convergent solver with an end-to-end prediction-error guarantee. In this paper, we study in-context kernel ridge regression (KRR) with Gaussian kernels and show that a standard softmax-attention transformer can approximate the KRR predictor during its forward pass by implementing preconditioned Richardson iteration on the associated kernel linear system. Under bounded-data assumptions, we construct a single-head transformer with

O(\log(1/ε))

blocks and MLP width

O(\sqrt{N/ε})

that achieves

ε

-accurate prediction for prompts of length

N

. Our construction reveals a functional decomposition within the transformer architecture: softmax attention produces a row-normalized Gaussian-kernel operator needed for cross-token interactions, while ReLU MLP layers act locally to approximate the intra-token scalar arithmetic required by the update. Empirically, we train GPT-2-style transformers on Gaussian-process regression tasks to further test the preconditioned Richardson interpretation. Through linear probing, we compare the transformer's layer-wise predictions with the step-wise outputs of classical KRR solvers and find that its error profiles align most consistently with preconditioned Richardson iteration. Ablation studies further support this interpretation. Together, our theory and experiments identify preconditioned Richardson iteration as a concrete mechanism that softmax-attention transformers can realize for nonlinear in-context Gaussian-kernel regression.

Mingsong Yan, Dongyang Li, Charles Kulick +1

Department of Mathematics, University of California, Santa Barbara, CA

Kernel Ridge Regression Transformer Attention Transformer Architectures Spectral Preconditioning Rectified Linear Unit Networks Newton

May 8, 2026cond-mat.dis-nn

Context-Gated Associative Retrieval: From Theory to Transformers

Hopfield networks and their generalizations have established deep connections among biological associative memories, statistical physics, and transformers. Yet most models treat retrieval as a fixed query-to-memory mapping, ignoring the role of external context in recall. In this work, we propose a two-stage associative memory architecture, wherein a context-gate subcircuit reshapes the retrieval energy landscape before and during recall. We show theoretically that context gating increases inter-memory separation while inducing sparsity, translating into exponential improvements in retrieval. Crucially, we prove that the system admits a unique self-consistent fixed point, revealing that the resulting retrieval state is driven by both a direct contextual bias and a second-order retrieval-gate feedback loop. We then bridge this theory to transformers; specifically, we evaluate a first-order approximation on Llama-3, confirming that in-context learning acts as context-gated retrieval. Native dynamics mirror our theory: context localizes a memory subspace, enabling the zero-shot query to cleanly discriminate. Ultimately, this framework provides a mechanistic link between associative memory theory and LLM phenomenology.

Moulik Choraria, Argyrios Gerogiannis, Vidhata Jayaraman +2

UIUC · University of Minnesota, Twin Cities · Stony Brook University

Dense Associative Memory In-Context Learning Transformer Architectures Theory Subspace

May 8, 2026cs.LG

When Attention Beats Fourier: Multi-Scale Transformers for PDE Solving on Irregular Domains

We study the problem of \emph{architecture selection} for deep learning models trained to solve partial differential equations (PDEs), asking when transformer-based architectures with learned attention outperform Fourier-domain neural operators. We introduce the \textbf{Multi-Scale Attention Transformer} (\msat{}), a deep learning architecture that encodes spatiotemporal solution histories as token sequences and trains end-to-end via a composite supervised objective with optional physics-informed regularization terms. We conduct a comprehensive empirical evaluation against nine baselines -- including physics-informed neural networks (PINNs), neural operators (FNO, DeepONet, GNOT), and state-space models (Mamba-NO) -- across five benchmark problems from the PINNacle suite, using identical train/test splits and reference data for all methods. \msat{} achieves state-of-the-art generalization on complex geometry problems (

L^2_\mathrm{rel} = 0.0101

on Heat2D-CG, a

3.7\times

improvement over FNO) at

34\,\mathrm{s}

total inference vs.\

120{,}812\,\mathrm{s}

for Mamba-NO. Ablation studies over the physics regularization component reveal a precise inductive bias tradeoff: physics priors reduce test error on diffusion-dominated problems but degrade generalization on chaotic and recirculating-flow regimes, directly characterizing the prior misspecification boundary. Approximation error bounds as a function of domain boundary complexity

κ

provide a theoretical basis for these empirical findings and a principled rule for architecture selection.

Brandon Yee, Pairie Koh, Jack Rodriguez +1

Physics Lab, Yee Collins Research Group

Partial Differential Equations Fourier Transformer Architectures Deep Learning Architectures Regularization

May 8, 2026cs.RO

NoiseGate: Learning Per-Latent Timestep Schedules as Information Gating in World Action Models

World Action Models (WAMs) are an emerging family of policies that tie robot action generation to future-observation modeling. In this work, we focus on the joint video--action modeling paradigm, where actions and imagined future observations are co-generated along a shared denoising or flow trajectory, so that perception, prediction, and control are coupled within one generative process. Existing WAMs typically realize this paradigm with a Mixture-of-Transformers (MoT), where video and action tokens interact through shared self-attention. This architecture can in principle assign a separate timestep

t_f

to each predicted latent frame, yet current systems collapse this degree of freedom onto a single shared scalar

t

. Under the noise-as-masking view of Diffusion Forcing, this shared schedule imposes the unjustified prior that every predicted latent is equally reliable for action generation. We instead view the per-latent schedule as a \emph{learnable information-gating policy}: by changing a latent frame's noise level, the policy modulates the reliability of its Key/Value contribution to the action tokens. We propose \textbf{NoiseGate}, which combines independent per-latent timestep sampling during backbone training, a lightweight Gating Policy Network that emits per-latent time increments during denoising, and task-reward optimization that trains the schedule policy without hand-crafted shape priors. Built on a joint video--action MoT backbone, NoiseGate delivers consistent gains on diverse RoboTwin random-scene manipulation tasks.

Wen Huang, Haoran Sun, Yongjian Guo +8

Tsinghua University · Peking University · JDT AI Infra +1

Efficient World-Action Model World Models Transformer Architectures

May 8, 2026cs.LG

FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast

SVD-based Low-rank compression reduces transformer parameters and nominal FLOPs, but these savings often translate poorly into real LLM serving speedups. We show that this gap is largely a runtime problem: factorized checkpoints fragment execution paths, and the resulting overhead differs substantially between prefill and autoregressive decode. We present FlashSVD v1.5, a unified inference runtime for serving SVD-compressed transformers. FlashSVD v1.5 maps diverse public SVD compression families to a common factorized representation and combines phase-specific kernels with dense-KV decode, packed MLP execution, and per-layer CUDA-graph replay to reorganize the low-rank serving path into a thin runtime. Across representative decoder-serving settings, FlashSVD v1.5 achieves up to 2.55x decode and 2.39x end-to-end speedup, and it attains 1.48x average decode and 1.44x average end-to-end speedup across multiple popular SVD compression families. These results suggest that practical low-rank acceleration requires runtime co-design, not compression algorithms alone. Our code is available at: https://github.com/Zishan-Shao/FlashSVD.

Wenhao Wu, Zishan Shao, Kangning Cui +5

Department of Electrical & Computer Engineering, Duke University · Department of Statistical Science, Duke University · Department of Computer Science, Wake Forest University +1

Singular Value Decomposition Large Language Model Compression Low-Rank Structure Decode Speedup Transformer Architectures

May 8, 2026cs.LG

Training-Induced Escape from Token Clustering in a Mean-Field Formulation of Transformers

Transformers perform inference by iteratively transforming token representations across layers. This layerwise computation has been studied empirically, and recent mean-field theories of Transformer dynamics explain how attention can drive token distributions toward clustering. However, existing mean-field analyses largely treat model parameters as prescribed, leaving open how training reshapes this clustering picture. We study this question in a noisy mean-field Transformer in which only a parameter-linear FFN is trained under

L^2

regularization. We find and analyze a training-induced phase in the dynamics: after initially following attention-driven clustering, the token distribution can leave the clustered regime near the final layers. Our mathematical analysis is based on an entropy-regularized interaction energy that captures the clustering bias of attention. More broadly, our results point toward a training-aware mean-field theory of Transformer dynamics, in which training and inference dynamics are treated together.

Noboru Isobe, Daisuke Inoue, Masaaki Imaizumi

1RIKEN Center for Advanced Intelligence Project · 2Imperial College London · 3The University of Tokyo +1

Transformer Attention Transformer Architectures Mean-Field Limit Inference Dynamics Clustering Escape

May 8, 2026cs.CL

Memory-Efficient Looped Transformer: Decoupling Compute from Memory in Looped Language Models

Recurrent LLM architectures have emerged as a promising approach for improving reasoning, as they enable multi-step computation in the embedding space without generating intermediate tokens. Models such as Ouro perform reasoning by iteratively updating internal representations while retaining a standard Key-Value (KV) cache across iterations, causing memory consumption to grow linearly with reasoning depth. Consequently, increasing the number of reasoning iterations can lead to prohibitive memory usage, limiting the practical scalability of such architectures. In this work, we propose Memory-Efficient Looped Transformer (MELT), a novel architecture that decouples reasoning depth from memory consumption. Instead of using a standard KV cache per layer and loop, MELT maintains a single KV cache per layer that is shared across reasoning loops. This cache is updated over time via a learnable gating mechanism. To enable stable and efficient training under this architecture, we propose to train MELT using chunk-wise training in a two phase procedure: interpolated transition, followed by attention-aligned distillation, both from the LoopLM starting model to MELT. Empirically, we show that MELT models fine-tuned from pretrained Ouro parameters outperform standard LLMs of comparable size, while maintaining a memory footprint comparable to those models and dramatically smaller than Ouro's. Overall, MELT achieves constant-memory iterative reasoning without sacrificing LoopLM performance, using only a lightweight post-training procedure.

Victor Conchello Vendrell, Arnau Padres Masdemont, Niccolò Grillo +3

Qualcomm AI Research is an initiative of Qualcomm Technologies, Inc.

Large Language Model Memory Recurrent Model Transformer Architectures

May 8, 2026cs.LO

Encoder-Decoder Transformers: Logical Characterizations and Periodicity

We give logical characterizations of encoder-decoder transformers, the foundational architecture for LLMs that also sees use in various settings that benefit from cross-attention, in the practical setting of floating-point numbers and soft attention. First, we characterize such transformers via a new temporal logic that extends propositional logic with a counting global modality over the encoder input and a past modality over the decoder input, as well as via a type of distributed automata. We consider three frameworks: with and without a final softmax step in the transformer, and in the setting where each model generates tokens via autoregression. Second, we show that both autoregressive transformers and sentences of counting propositional logic - the fragment of the previous logic obtained by omitting the past modality - recognize exactly the commutative star-free languages. Finally, we find that the sequences of tokens the transformers generate are ultimately periodic (and each token appears in the period at most once). This allows us to characterize autoregressive transformers via sentences of counting propositional logic that generate tokens without autoregression, i.e., we can effectively eliminate recursion from the transformers.

Veeti Ahvonen, Damian Heiman, Antti Kuusisto +2

Mathematics Research Centre, Tampere University, Tampere, Finland

Transformer Encoder Transformer Architectures Large Language Model Decoding Cross-Attention Automata Modalities

May 8, 2026cs.LG

Revisiting Transformer Layer Parameterization Through Causal Energy Minimization

Transformer blocks typically combine multi-head attention (MHA) for token mixing with gated MLPs for token-wise feature transformation, yet many choices in their parameterization remain largely empirical. We introduce Causal Energy Minimization (CEM), a framework that recasts Transformer layers as optimization steps on conditional energy functions while explicitly accounting for layer parameterization. Extending prior energy-based interpretations of attention, CEM shows that weight-tied MHA can be derived as a gradient update on an interaction energy, and that a gated MLP with shared up/down projections can be viewed through an element-wise energy. This perspective identifies a design space for Transformer layers that includes within-layer weight sharing, diagonal-plus-low-rank interactions, lightweight preconditioners, and recursive updates. We evaluate CEM-derived layers in language-modeling experiments at the moderate hundred-million-parameter scale. Despite their constrained parameterizations, these layers train stably and can match corresponding Transformer baselines. Overall, our results suggest that CEM provides a useful lens for understanding Transformer layer parameterization, connecting Transformer architectures to energy-based models and motivating further exploration of energy-guided layer designs.

Jin Xu, Camille Couturier, Victor Rühle +2

Microsoft · Microsoft Research Cambridge

Transformer Architectures Multi-Head Attention Free Energy Principle Causal

May 8, 2026cs.CV

Dynamic Mode Decomposition along Depth in Vision Transformers

Recent work has shown that contiguous vision transformer (ViT) blocks (a) can be replaced by a linear map and (b) organize into recurrent phases of computation. We ask whether these observations coincide: does ViT depth implement approximately \textit{autonomous linear} dynamics, admitting a single operator

K

applied recurrently across a contiguous span? We test this using Dynamic Mode Decomposition (DMD), which fits

K

from selected, consecutive hidden-state pairs and predicts

p

steps ahead via

K^p

. On four pretrained DINO ViTs, we study the regularization, rank, and calibration budget required for stable fitting. For short spans (

p \leq 4

),

K^p

tracks an unconstrained endpoint map to within

0.02

cosine similarity on DINOv3-H/16+, while also recovering intermediate activations at each skipped block. At early cut starts, the fitted operators compress to rank

\ll d

with minimal calibration data, and across tokens, \texttt{cls} is most amenable to linearization; both properties decay monotonically with depth. Yet this local fidelity does not transfer downstream. At the final hidden state, after propagating through the remaining blocks, an identity baseline becomes competitive.

Nishant Suresh Aswani, Saif Eddin Jabari

NYU Tandon & NYU Abu Dhabi

Self-Supervised Vision Transformers Vision Transformer Dynamic Mode Decomposition Transformer Architectures Hidden States Recurrence

May 8, 2026cs.LG

Approximation Error Upper and Lower Bounds for Hölder Class with Transformers

We explore the expressive power of Transformers by establishing precise approximation error upper and lower bounds for Hölder class. Specifically, a new approximation upper bound is derived for the standard Transformer architecture equipped with Softmax operators, ReLU activation functions, and residual connections. We prove that a Transformer network composed of at most

\mathcal{O}(\varepsilon^{-{d_{0}}/α})

blocks can approximate any bounded Hölder function with

d_{0}

-dimensional input and smoothness

α\in(0,1]

under any accuracy

\varepsilon>0

. In the case of approximation lower bounds, leveraging the VC-dimension upper bound, we are the first to rigorously prove that Transformers demand for at least

Ω(\varepsilon^{-{d_{0}}/({4α})})

blocks to achieve the

\varepsilon

approximation accuracy. As a final step, we extend the derived results for standard Transformers to a general regression task and establish the corresponding excess risk rates demonstrating Transformers' empirical effectiveness in real-world settings.

Xin He, Yuling Jiao, Xiliang Lu +1

School of Mathematics and Statistics, Wuhan University, Wuhan, China · School of Artificial Intelligence, Wuhan University, Wuhan, China · Hubei Key Laboratory of Computational Science, Wuhan University, Wuhan, China

Transformer Architectures Approximation Generalization Bounds Lower Bounds Intrinsic Dimensionality Expressivity

May 8, 2026cs.LG

Beyond Linear Attention: Softmax Transformers Implement In-Context Reinforcement Learning

In-context reinforcement learning (ICRL) studies agents that, after pretraining, adapt to new tasks by conditioning on additional context without parameter updates. Existing theoretical analyses of ICRL largely rely on linear attention, which replaces the softmax function in the standard attention with an identity mapping. This paper provides the first theoretical understanding of ICRL without making the unrealistic linear attention simplification. In particular, we consider the standard softmax attention used in practice. We show that, with certain parameters, the layerwise forward pass of a Transformer with such softmax attention is equivalent to iterative updates of a weighted softmax temporal difference (TD) learning algorithm. Here, weighted softmax TD is a new RL algorithm that performs policy evaluation in kernel space and adopts both linear TD and tabular TD as special cases. We also prove that under a certain contraction condition, the policy evaluation error decays as the number of layers grows, with the identified parameters above. Finally, we prove that those parameters are a global minimizer of a pretraining loss, explaining their emergence in our numerical experiments.

Zixuan Xie, Xinyu Liu, Claire Chen +3

University of Virginia · California Institute of Technology · Purdue University

In-Context Learning Linear Attention Transformer Architectures Gumbel-Softmax Relaxation Layer-Wise Policy Evaluation

May 8, 2026cs.CV

Amortized-Precision Quantization for Early-Exit Vision Transformers

Vision Transformers (ViTs) achieve strong performance across vision tasks, yet their deployment with low-precision early exiting remains fragile. Existing quantization methods assume static full-depth execution, making them unstable when exit decisions are perturbed by quantization noise, which can amplify errors along dynamic inference paths. In this paper, we introduce Amortized-Precision Quantization (APQ), a utilization-aware formulation that accounts for layer-wise stochastic exposure to quantization noise and reveals depth-precision trade-offs. Building on APQ, we propose Mutual Adaptive Quantization with Early Exiting (MAQEE), a bi-level framework that jointly optimizes exit thresholds and bit-widths under explicit risk control to improve inference stability. MAQEE establishes a superior Pareto frontier in the accuracy-efficiency trade-off, reducing BOPs by up to 95% while maintaining accuracy and outperforming strong baselines by up to 20% across classification, detection, and segmentation tasks.

Rui Fang, Hsi-Wen Chen, Ming-Syan Chen

National Taiwan University

Self-Supervised Vision Transformers Vision Transformer Layer-Wise Transformer Architectures

May 8, 2026stat.ML

Spectrum-Adaptive Generalization Bounds for Trained Deep Transformers

Understanding why trained Transformers generalize well is a fundamental problem in modern machine learning theory, and complexity-based generalization bounds provide a principled way to study this question. While existing norm-based bounds for Transformers remove the explicit polynomial dependence on the hidden dimension, they typically impose fixed norm constraints specified a priori and can exhibit unfavorable exponential dependence on depth. In this paper, we derive spectrum-adaptive post hoc generalization bounds for multi-layer Transformers. Under layerwise spectral norm control, the bounds are expressed in terms of layerwise Schatten quantities of the query-key, value, and feedforward weight matrices. Since the Schatten indices need not be fixed a priori and can instead be selected after training, separately for each matrix type and layer, the bounds adaptively trade off spectral complexity against the dimension- and depth-dependent factors according to the learned singular-value profiles. Empirical comparisons of BERT-adapted proxies for the leading complexity factors suggest that the proxies induced by our bounds grow more slowly with depth and hidden dimension than the corresponding norm-based proxies. Overall, our results provide a complexity-based perspective on how the spectral structure of trained Transformers is reflected in generalization analyses.

Mana Sakai, Masaaki Imaizumi

The University of Tokyo · RIKEN Center for Advanced Intelligence Project · Kyoto University

Transformer Architectures Generalization Bounds

May 8, 2026cs.CV

TAS-LoRA: Transformer Architecture Search with Mixture-of-LoRA Experts

Transformer architecture search (TAS) discovers optimal vision transformer (ViT) architectures automatically, reducing human effort to manually design ViTs. However, existing TAS methods suffer from the feature collapse problem, where subnets within a supernet fail to learn subnet-specific features, mainly due to the shared weights in a supernet, limiting the performance of individual subnets. To address this, we propose TAS-LoRA, a novel method that introduces parameter-efficient low-rank adaptation (LoRA) to enable subnet-specific feature learning, while maintaining computational efficiency. TAS-LoRA incorporates a Mixture-of-LoRAExperts (MoLE) strategy, where a lightweight router dynamically assigns LoRA experts based on subnet architectures, and introduces a group-wise router initialization technique to encourage diverse feature learning across experts early in training. Extensive experiments on ImageNet and several transfer learning benchmarks, including CIFAR-10/100, Flowers, CARS, and INAT-19, demonstrate that TAS-LoRA mitigates feature collapse effectively, improving performance over state-of-the-art TAS methods significantly.

Jeimin Jeon, Hyunju Lee, Bumsub Ham

Yonsei University · Articron Inc. · Korea Institute of Science and Technology (KIST)

Self-Supervised Vision Transformers Hip-Lora Transformer Architectures Subspace Experts

May 8, 2026cs.LG

Geometric Asymmetry in MoE Specialization: Functional Decorrelation and Representational Overlap

Mixture-of-Experts (MoE) architectures achieve scalable capacity through sparse routing, yet the geometric structure of expert specialization remains poorly understood. We introduce a unified Jacobian-PCA-Grassmann framework for analyzing MoE layers in both function space and representation space. Across pretrained MoE Transformers (Mistral, Qwen), we find a consistent structural asymmetry: experts exhibit strong functional decorrelation (consistently low, near-zero cross-expert Jacobian alignment) while their routed representations occupy distinct but partially overlapping subspaces. This indicates that functional decorrelation and representation overlap coexist rather than coincide in MoE specialization. Controlled routing experiments further indicate that routing sparsity appears to be a key factor shaping this geometry: top-k routing induces sharper functional separation and larger subspace divergence, whereas fully soft routing yields more entangled expert structure. Together, these results suggest a geometric interpretation in which MoE layers may be viewed as implementing locally decorrelated operators over overlapping submanifolds on a shared representation manifold, and provide a general diagnostic framework for studying conditional computation in modern Transformer architectures.

Feilong Liu

Mixture-Of-Experts Layer-Wise Transformer Architectures Asymmetry Jacobian

May 8, 2026cs.CV

Attention Transfer Is Not Universally Effective for Vision Transformers

A recent work shows that Attention Transfer, which transfers only the attention patterns from a pre-trained teacher Vision Transformer (ViT) to a randomly initialized standard student ViT, is sufficient to recover the full benefit of the teacher's pre-trained weights. We revisit this finding on a comprehensive benchmark of 20 teachers from 11 well-known ViT families and reveal that Attention Transfer is not universally effective. While 7 families transfer successfully, 4 consistently fail, falling up to 5.1% below the from-scratch no-transfer baseline. Further results demonstrate that this failure is family-consistent across model sizes, and persists under extended training durations, different transfer datasets, and out-of-distribution evaluations. Controlled analyses then consistently localize the problem to the attention-routing channel, indicating that the key issue is not whether the student can match the teacher's attention patterns, but whether the matched patterns remain functional for the student. Crucially, we identify architectural mismatch between the pre-trained teacher and the standard student as the primary mechanism. By adding only the teacher's native architectural components to the student in a randomly initialized state, we completely reverse the failure for all 4 families. Notably, these components alone do not improve from-scratch training, confirming that they specifically unlock the usability of the teacher's attention. We further systematically show that this failure is not explained by the inadequate choice of transfer loss or by differences in pre-training recipes. Our findings refine the prevailing understanding of attention in ViT representations: attention is sufficient \textit{only} when the student architecture matches the teacher.

Huaiyuan Qin, Muli Yang, Gabriel James Goenawan +4

Institute for Infocomm Research (I2R), A*STAR, Singapore · 3Shanghai Jiao Tong University · 2Sichuan University

Self-Supervised Vision Transformers Vision Transformer Attention Layers Transformer Architectures

May 8, 2026cs.CV

SAFformer:Improving Spiking Transformer via Active Predictive Filtering

Spiking Neural Networks (SNNs) offer notable advantages in biological plausibility and energy efficiency, making them promising candidates for building low-power Transformers. However, existing Spiking Transformers largely adhere to a passive reactive paradigm, which struggles to focus on task-relevant information and incurs substantial computational overhead when processing redundant visual data. To overcome this fundamental yet underexplored limitation, we propose SAFformer, a novel Spiking Transformer architecture based on an active predictive filtering paradigm. Inspired by the brain's predictive coding mechanism, SAFformer actively suppresses predictable signals and focuses on salient visual features. Extensive experiments show that SAFformer establishes new state-of-the-art performance on CIFAR-10/100 and CIFAR10-DVS. Remarkably, on ImageNet-1K, it achieves 80.44% Top-1 accuracy with only 26.58M parameters and an energy consumption of 5.88 mJ, demonstrating an exceptional balance between accuracy and efficiency.

Zequan Xie, Weiming Zeng, Yunhua Chen +3

School of Computer Science and Technology, Guangdong University of Technology, Guangzhou, China · Faculty of Science, Hong Kong Baptist University, China · School of Electronic Information, Wuhan University, China

Spiking Neural Networks Transformer Architectures Bayesian Filtering

May 8, 2026cs.LG

When Symbol Names Should Not Matter: A Logistic Theory of Fresh-Symbol Classification

Template tasks have emerged as a clean testbed for asking whether transformers reason with abstract symbols rather than concrete token names. We study the fixed-label classification version of this problem, where train and test examples share latent templates but may use disjoint vocabularies. Unlike next-token prediction, the model need not emit unseen symbols; it must learn a decision rule invariant to symbol renaming. We analyze regularized kernel logistic classification in the transformer-kernel regime. Our main result decomposes the learned predictor into an ideal template-level classifier and a finite-sample perturbation caused by accidental token overlaps in the training data. We encode these overlaps by a colored collision graph and prove high-probability margin-transfer guarantees for fresh-symbol classification. This perspective extends template-based analyses to logistic classification and refines scalar diversity conditions: vocabulary size controls the average rate of collisions, but collision geometry controls whether the ideal classification margin is preserved. More broadly, the same perturbation framework applies to abstraction-augmented inputs, yielding a general margin-versus-collision criterion for identifying when prompting strategies improve fresh-symbol generalization. Synthetic template experiments illustrate the predicted roles of regularization, sample size, and transformer-kernel structure.

Wenjie Guan, Jelena Bradic

Department of Statistics and Data Science, Cornell University

Symbol-Transformer Architectures Training-Inference Mismatch Next-Token Prediction Latent Variable

May 8, 2026cs.LG

LoopQ: Quantization for Recursive Transformers

Looped language models (LoopLMs) improve parameter efficiency by recursively reusing Transformer blocks, enabling deeper computation under a fixed model size. However, this reuse makes LoopLMs more fragile under post-training quantization (PTQ). We present the first systematic study of quantization in LoopLMs and identify three challenges: distribution shift across roles, state reuse across loop transitions, and recursive error accumulation. To address these challenges, we propose LoopQ, a loop-aware PTQ framework that preserves a shared quantized backbone while introducing lightweight adaptations. LoopQ combines activation scaling, selective transformation, cross-loop state alignment, and trajectory-aware optimization to reduce distributional mismatch within loops and error accumulation across loops. Experiments across seven benchmarks show that, under W4A4 quantization, LoopQ improves average downstream accuracy by 68.8% and reduces average perplexity by 87.7% compared with the strongest static PTQ baseline.

Rui Fang, Hsi-Wen Chen, Ming-Syan Chen

National Taiwan University

Post-Training Quantization Transformer Architectures

May 7, 2026cs.CV

Not All Tokens Need 40 Steps: Heterogeneous Step Allocation in Diffusion Transformers for Efficient Video Generation

Diffusion Transformers (DiTs) have achieved state-of-the-art video generation quality, but they incur immense computational cost because standard inference applies the same number of denoising steps uniformly to every token in the sequence. It is well known that human vision ignores vast amounts of redundant motion. Why, then, do our densest models treat every spatiotemporal token with equal priority? In this paper, we introduce Heterogeneous Step Allocation (HSA), a training-free inference algorithm that assigns varying step budgets to different spatiotemporal tokens based on their velocity dynamics. To resolve the resulting sequence-length mismatch without sacrificing global context, HSA introduces a KV-cache synchronization mechanism that allows active tokens to attend to the full sequence while entirely bypassing inactive tokens. Furthermore, we derive a cached Euler update that advances the latent states of skipped tokens in a single operation without additional model evaluations. We evaluate HSA on the Wan-2 and LTX-2 models for both text-to-video (T2V) and image-to-video (I2V) generation. Our results demonstrate that HSA significantly outperforms previous state-of-the-art caching methods and the vanilla Flow Matching baseline, especially at aggressive acceleration regimes (e.g., 50% and 25% runtimes). Crucially, HSA achieves a superior quality-runtime Pareto frontier without the need for expensive offline profiling, robustly preserving structural integrity and generation quality even under tight computational budgets. Project page: https://ernestchu.github.io/hsa

Ernie Chu, Vishal M. Patel

Johns Hopkins University Baltimore, MD 21218

Diffusion Transformers Spatial Tokens Transformer Architectures Redundancy

May 7, 2026cs.LG

ProtSent: Protein Sentence Transformers

Protein language models (pLMs) produce per-residue representations that capture evolutionary and structural information, yet their mean-pooled sequence embeddings are not explicitly trained to reflect functional, evolutionary or structural similarity between proteins. We present Protein Sentence Transformers (ProtSent), a contrastive fine-tuning framework for adapting PLMs into general-purpose embedding models. ProtSent trains with MultipleNegativesRankingLoss across five protein-pair datasets: Pfam families, structurally derived hard negatives, AlphaFold DB structural pairs, and StringDB protein--protein interactions, and Deep Mutational Scanning data. We evaluate on 23~downstream tasks using frozen embeddings with a k-nearest-neighbor probe to measure embedding neighborhood quality. On ESM-2 150M, ProtSent improves 15 of 23 tasks, with gains of +105% on remote homology detection, +17% on variant effect prediction, and +19.9% Recall@1 on SCOPe-40 structural retrieval. The 35M variant improves 16 of 23 tasks with +40.5% on remote homology and +15.5% Recall@1 on SCOPe-40. Contrastive fine-tuning restructures the embedding space to better capture protein function and structure, without any task-specific supervision. We release the models, public data, and training recipe and code.

Dan Ofer, Oriel Perets, Michal Linial +1

Department of Computer and Information Science Ben-Gurion University of the Negev

Language Modeling Transformer Architectures

May 7, 2026cs.LG

Transformers Efficiently Perform In-Context Logistic Regression via Normalized Gradient Descent

Transformers have demonstrated remarkable in-context learning (ICL) capabilities. The strong ICL performance of transformers is commonly believed to arise from their ability to implicitly execute certain algorithms on the context, thereby enhancing prediction and generation. In this work, we investigate how transformers with softmax attention perform in-context learning on linear classification data. We first construct a class of multi-layer transformers that can perform in-context logistic regression, with each layer exactly performing one step of normalized gradient descent on an in-context loss. Then, we show that our constructed transformer can be obtained through (i) training a single self-attention layer supervised by one-step gradient descent, and (ii) recurrently applying the trained layer to obtain a looped model. Training convergence guarantees of the self-attention layer and out-of-distribution generalization guarantees of the looped model are provided. Our results advance the theoretical understanding of ICL mechanism by showcasing how softmax transformers can effectively act as in-context learners.

Chenyang Zhang, Yuan Cao

School of Computing & Data Science, The University of Hong Kong

In-Context Learning Transformer Architectures In-Context Linear Attention Logistic Regression Gradient Descent

May 7, 2026cs.LG

Is One Layer Enough? Understanding Inference Dynamics in Tabular Foundation Models

Transformer-based tabular foundation models (TFMs) dominate small to medium tabular predictive benchmark tasks, yet their inference mechanisms remain largely unexplored. We present the first large-scale mechanistic study of layerwise dynamics in 6 state-of-the-art tabular in-context learning models. We explore how predictions emerge across depth, identify distinct stages of inference and reveal latent-space dynamics that differ from those of language models. Our findings indicate substantial depthwise redundancy across multiple models, suggesting iterative refinement with overlapping computations during inference stages. Guided by these insights, we design a proof-of-concept, looped single-layer model that uses only 20% of the original model's parameters while achieving comparable performance. The code is available at https://github.com/amirbalef/is_one_layer_enough.

Amir Rezaei Balef, Mykhailo Koshil, Katharina Eggensperger

TU Dortmund University, Dortmund, Germany · Lamarr Institute for Machine Learning and Artificial Intelligence, Dortmund, Germany · University of Tübingen, Tübingen, Germany

Tabular Foundation Models Transformer Architectures Layer-Wise Foundation Model Inference Dynamics Latent Dynamics

May 7, 2026cs.LG

Cubit: Token Mixer with Kernel Ridge Regression

Since its introduction in 2017, the Transformer has become one of the most widely adopted architectures in modern deep learning. Despite extensive efforts to improve positional encoding, attention mechanisms, and feed-forward networks, the core token-mixing mechanism in Transformers remains attention. In this work, we show that the attention module in Transformers can be interpreted as performing Nadaraya-Watson regression, where it computes similarities between tokens and aggregates the corresponding values accordingly. Motivated by this perspective, we propose Cubit, a potential next-generation architecture that leverages Kernel Ridge Regression (KRR), while the vanilla Transformer relies on Nadaraya-Watson regression. Specifically, Cubit modifies the classical attention computation by incorporating the closed-form solution of KRR, combining value aggregation through kernel similarities with normalization via the inverse of the kernel matrix. To improve the training stability, we further propose the Limited-Range Rescale (LRR), which rescales the value layer within a controlled range. We argue that Cubit, as a KRR-based architecture, provides a stronger mathematical foundation than the vanilla Transformer, whose attention mechanism corresponds to Nadaraya-Watson regression. We validate this claim through comprehensive experiments. The experimental results suggest that Cubit may exhibit stronger long-sequence modeling capability. In particular, its performance gain over the Transformer appears to increase as the training sequence length grows.

Chuanyang Zheng, Jiankai Sun, Yihang Gao +6

Transformer Attention Transformer Architectures Sequence Modeling Kernel Ridge Regression Ridge

May 7, 2026cs.AI

Patch-Effect Graph Kernels for LLM Interpretability

Mechanistic interpretability aims to reverse-engineer transformer computations by identifying causal circuits through activation patching. However, scaling these interventions across diverse prompts and task families produces high-dimensional, unstructured datasets that are difficult to compare systematically. We propose a framework that reframes mechanistic analysis as a graph machine-learning problem by representing activation-patching profiles as patch-effect graphs over model components. We introduce three graph-construction methods: direct-influence via causal mediation, partial-correlation, and co-influence and apply graph kernels to analyze the resulting structures. Evaluating this approach on GPT-2 Small using Indirect Object Identification (IOI) and related tasks, we find that patch-effect graphs preserve discriminative structural signals. Specifically, localized edge-slot features provide higher classification accuracy than global graph-shape descriptors. A screened paired-patching validation suggests that CI and PC selected candidate edges correspond to stronger activation-influence effects than random or low-rank candidates. Crucially, by evaluating these representations against rigorous prompt-only and raw patch-effect controls, we make the evidential scope of the benchmark explicit: graph features compress structured patching signal, while raw tensors and surface cues define strong baselines that any circuit-level claim should address. Ultimately, our framework provides a compression and evaluation pipeline for comparing patching-derived structures under controlled baselines, separating robust slice-discriminative evidence from stronger task-general causal-circuit claims.

Ruben Fernandez-Boullon, David N. Olivieri

Department of Computer Science, University of Vigo, Ourense 32004, Spain

Mechanistic Interpretability Transformer Architectures

May 7, 2026cs.LG

Federation of Experts: Communication Efficient Distributed Inference for Large Language Models

Mixture of experts has emerged as the primary mechanism for making Large Language Models (LLMs) computationally efficient. However, in distributed settings, communicating token embeddings between experts is a significant bottleneck. We present the novel Federation of Experts (FoE) architecture. FoE restructures the MoE block of a transformer layer into multiple MoE clusters. Each cluster is responsible for only one of the KV heads and expert parallelism is applied between those experts. Between clusters, a sum synchronizes the post-attention residuals, which then drives routing and dispatch for the next MoE block. In a single-node setting, FoE completely eliminates all-to-all communication as all experts within a group are contained on the same GPU. In multi-node settings, FoE confines all-to-all communication to the intra-node fabric, thus significantly reducing communication overhead. An implementation of FoE finds that on LongBench, FoE significantly improves inference throughput and latency in both single-node and multi-node settings, reducing the end-to-end forward-pass latency by up to 5.2x, TTFT by 3.62x, and TBT by 1.95x. It does so while achieving comparable generation quality to a mixture of experts model of the same size and training configuration.

Muhammad Shahir Abdurrahman, Chun Deng, Azalia Mirhoseini +1

Stanford University

Transformer Architectures Token Embeddings Single Pass

May 7, 2026cs.CV

Dynamic Pondering Sparsity-aware Mixture-of-Experts Transformer for Event Stream based Visual Object Tracking

Despite significant progress, RGB-based trackers remain vulnerable to challenging imaging conditions, such as low illumination and fast motion. Event cameras offer a promising alternative by asynchronously capturing pixel-wise brightness changes, providing high dynamic range and high temporal resolution. However, existing event-based trackers often neglect the intrinsic spatial sparsity and temporal density of event data, while relying on a single fixed temporal-window sampling strategy that is suboptimal under varying motion dynamics. In this paper, we propose an event sparsity-aware tracking framework that explicitly models event-density variations across multiple temporal scales. Specifically, the proposed framework progressively injects sparse, medium-density, and dense event search regions into a three-stage Vision Transformer backbone, enabling hierarchical multi-density feature learning. Furthermore, we introduce a sparsity-aware Mixture-of-Experts module to encourage expert specialization under different sparsity patterns, and design a dynamic pondering strategy to adaptively adjust the inference depth according to tracking difficulty. Extensive experiments on FE240hz, COESOT, and EventVOT demonstrate that the proposed approach achieves a favorable trade-off between tracking accuracy and computational efficiency. The source code will be released on https://github.com/Event-AHU/OpenEvTracking.

Shiao Wang, Xiao Wang, Duoqing Yang +5

School of Computer Science and Technology, Anhui University, Hefei 230601, China · Beijing Institute of Technology, Beijing, China · Peng Cheng Laboratory, Shenzhen, China +2

Multi-Object Tracking Self-Supervised Vision Transformers Sparsity Transformer Architectures

May 7, 2026cs.LG

Beyond Autoregressive RTG: Conditioning via Injection Outside Sequential Modeling in Decision Transformer

Decision Transformer (DT) formulates offline reinforcement learning as autoregressive sequence modeling, achieving promising results by predicting actions from a sequence of Return-to-Go (RTG), state, and action tokens. However, RTG is a scalar that summarizes future rewards, containing far less information than typical state or action vectors, yet it consumes the same computational budget per token. Worse, the self-attention cost of Transformers grows quadratically with sequence length, so including RTG as a separate token adds unnecessary overhead. We propose SlimDT, which removes RTG from the autoregressive sequence. Instead, we inject RTG information into the state representations before the sequential modeling step, allowing the Transformer to process only a compact (state, action) sequence. This reduces the sequence length by one-third, directly improving inference efficiency. On the D4RL benchmark, SlimDT surpasses standard DT across various tasks and achieves performance comparable to existing state-of-the-art methods. Decoupling a sparse conditioning signal from an information-rich sequence thus yields both computational gains and higher task performance.

Yongyi Wang, Hanyu Liu, Lingfeng Li +6

School of Computer Science Peking University Beijing, China 100871

Transformer Architectures Sequence Modeling Offline Reinforcement Learning Autoregressive Model

May 7, 2026cs.LG

The E $Δ$ -MHC-Geo Transformer: Adaptive Geodesic Operations with Guaranteed Orthogonality

We present the E

Δ

-MHC-Geo Transformer, a novel architecture that unifies Manifold-Constrained Hyper-Connections (mHC), Deep Delta Learning (DDL), and the Cayley transform to obtain input-adaptive, unconditionally orthogonal residual connections. Unlike DDL, whose Householder operator is orthogonal only at

β\in \{0,2\}

, our Data-Dependent Cayley rotation

Q(x)=(I+(β/2)A(x))^{-1}(I-(β/2)A(x))

preserves orthogonality for all

β

and all inputs. To handle negation, an eigenvalue

-1

case that Cayley provably excludes, we introduce the E

Δ

-MHC-Geo Hybrid, which combines Cayley rotation with Householder reflection via a learned operator-selection gate

X'=γ(X)Q(X)X+(1-γ(X))H_2(X)X

. A midpoint-collapse regularizer,

4γ(1-γ)

, encourages boundary gate decisions, where each selected component is orthogonal. In matched-parameter comparisons, with approximately 1.79M parameters per model and mean +/- standard deviation over 3 seeds, against four baselines including the concurrent JPmHC, E

Δ

-MHC-Geo achieves the best long-horizon stability, 1.9x over JPmHC and 3.8x over GPT; the best near-

π

rotation loss, 4.5x over JPmHC on single-plane; strong norm preservation, with 0.001 mean deviation; and 0.96 negation cosine alignment in a diagnostic reflection probe, all with 33% fewer layers. While JPmHC's wider representation excels on pure rotation, its finite Cayley residual mixer excludes an exact

λ=-1

operator and has no reflection branch, motivating our hybrid approach for accessing both connected components of

O(n)

.

Arash Shahmansoori

Independent Researcher

Residual Connections Orthogonality Transformer Architectures Geodesic Antibody

May 7, 2026cs.LG

Transformer-Based Wildlife Species Classification from Daily Movement Trajectories

Inferring the identity of wildlife species from daily movement data alone is a challenging task. We train sequence models on large-scale, 7-species GPS trajectories from the Movebank platform. Trajectories models are evaluated using a protocol in which entire telemetry studies or regions are heldout during testing. We compare Transformer-based sequence models to LSTM, CNN, and Temporal Convolutional Networks, and find that Transformers consistently achieve higher balanced accuracy with gains of approximately 8 to 22 percentage points, depending on the species and experimental setting. In an elephant binary classification task with 1-hour resolution, the Transformer achieves a balanced accuracy of 0.83 and an AUC of 0.92, substantially outperforming all baseline models. We examine, under data-limited conditions, feature representations by analyzing the differences between a basic displacement-based encoding and an expanded range of movement descriptors that include speed, direction, and turning behavior. With feature augmentation, we see clear performance gains, especially for underrepresented and sparsely represented species, such as large carnivores, lions, and Zebras. Finally, experiments comparing 1-hour and 30-minutetemporal resolutions show that while finer sampling can capture short-term movement patterns for some species, a unified 1-hour resolution yields more promising performance across studies by reducing missing data and ensuring consistent temporal coverage.

Obed Irakoze, Prasenjit Mitra

Department of Electrical & Computer Engineering Carnegie Mellon University Africa Kigali, Rwanda

Species Identification Species Temporal Convolutional Networks Transformer Architectures

May 7, 2026cs.LG

Training Transformers for KV Cache Compressibility

Long-context language modeling is increasingly constrained by the Key-Value (KV) cache, whose memory and decode-time access costs scale linearly with the prefix length. This bottleneck has motivated a range of context-compression methods, from token-level summarization to recent optimization-based KV compression methods. These post-hoc methods operate on the KV cache of a fixed pretrained model, so their effectiveness is fundamentally limited by how well the model's internal representations can be compressed. In this work, we formalize the notion of KV compressibility and show that it is a property of the learned representations, rather than of the context alone. We prove that almost any sequence-to-vector function admits both highly compressible and inherently non-compressible transformer implementations, highlighting the need to guide transformers toward compressible representations during training. Motivated by this, we propose KV-Compression Aware Training (KV-CAT), a continued pretraining procedure that incentivizes the emergence of compressible representations. We introduce a train-time KV sparsification policy that masks KV slots during training. This forces the model to use fewer KV slots and encourages it to learn representations amenable to post-hoc compression. Empirically, we show that KV-CAT improves the quality-budget tradeoff of downstream compression methods across retrieval, long-context question answering, and perplexity-based evaluation of compressed-prefix continuation.

Yoav Gelberg, Yam Eitan, Michael Bronstein +2

University of Oxford · 2Technion – Israel Institute of Technology · 3AITHYRA +1

Key-Value Cache Compression Efficient Long-Context Inference Key-Value Cache Transformer Architectures Cache Sequence Modeling+1

May 7, 2026stat.ML

Transformers Provably Implement In-Context Reinforcement Learning with Policy Improvement

We investigate the ability of transformers to perform in-context reinforcement learning (ICRL), where a model must infer and execute learning algorithms from trajectory data without parameter updates. We show that a linear self-attention transformer block can provably implement policy-improvement methods, including semi-gradient SARSA and actor-critic, via explicit parameter constructions. Beyond existence, we design a teacher-mimicking training procedure, analyze its gradient-flow dynamics, and establish the first convergence guarantee in the ICRL literature: under suitable richness conditions on the training MDP distribution, gradient flow converges locally and exponentially to an optimal parameter manifold corresponding to the desired RL update. Empirically, training transformers on randomly generated tabular MDPs confirms these predictions: the learned models recover the parameter structure of our explicit constructions and, when deployed on unseen MDPs, deliver strong in-context control performance. Together, these results illuminate how transformer architectures internalize and execute classical reinforcement learning algorithms in context, bridging mechanistic understanding and training dynamics in ICRL.

Haodong Liang, Lifeng Lai

Department of Electrical and Computer Engineering, University of California, Davis

Offline Reinforcement Learning In-Context Learning Transformer Architectures Transformer Attention

May 7, 2026cs.AI

HyperLens: Quantifying Cognitive Effort in LLMs with Fine-grained Confidence Trajectory

While Large Language Models (LLMs) achieve strong performance across diverse tasks, their inference dynamics remain poorly understood because of the limited resolution of existing analysis tools. In this work, we identify an intrinsic magnification mechanism in transformer architectures: deeper layers inherently magnify the small changes of layer-wise confidence, providing a fine-grained confidence trajectory. Building on this insight, we introduce HyperLens, a high-resolution probe designed to trace confidence trajectories and quantify the cognitive effort during inference. Across LLMs and datasets, HyperLens reveals a consistent divergence in confidence trajectories that separates complex from simple tasks. We abstract this pattern into a quantitative cognitive effort metric. Our analysis reveals a fundamental principle: complex tasks consistently require higher cognitive effort. Finally, we provide a mechanistic diagnosis of a common side effect of standard Supervised Fine-Tuning (SFT): it can reduce cognitive effort and consequently degrade performance on in-domain tasks.

Chengda Lu, Xiaoyu Fan, Wei Xu

1IIIS, Tsinghua University · 2Shanghai Qi Zhi Institute.

LLM Inference Optimization Cognitive Load Cognitive Science Confidence Estimation Transformer Architectures

May 7, 2026cs.LG

Budgeted Attention Allocation: Cost-Conditioned Compute Control for Efficient Transformers

Transformers usually expose one inference cost per trained model, while deployed systems often need multiple cost-quality operating points. We study Budgeted Attention Allocation, a monotone head-gating mechanism conditioned on a requested attention budget. Dense warm-starting is important for stability: on a robust synthetic sequence task, one budgeted model reaches 99.7% accuracy at 0.303 estimated attention cost and 100.0% accuracy at 0.504 cost. On held-out AG News with a custom word-level transformer, hard-gate adaptation turns soft cost control into measured single-thread CPU speed, reaching 82.1% accuracy with 1.28x speedup at budget 0.50. In pretrained BERT-Mini AG News, budgeted structural pruning reaches 87.6% accuracy with 1.20x speedup at budget 0.50; a validation-ranked zero-shot dense post-hoc structural baseline reaches 86.1%, and one recovery epoch raises that per-budget specialist to 87.9%. On DBpedia14, BERT-Mini budgeted gates reach 97.4% at exact budget 0.50 versus 96.6% for dense full attention. Static fixed-budget gates and recovered dense specialists remain strong. The contribution is therefore not universal dominance, but a reproducible feasibility study of one controllable checkpoint across budgets that can trade attention cost for accuracy and be converted into measured structural speedups on small CPU benchmarks.

Amrit Nidhi

Independent Researcher

Transformer Attention Transformer Architectures Token Budget Allocation Inference Cost

May 7, 2026cs.LG

Distributional Spectral Diagnostics for Localizing Grokking Transitions

In grokking, a model first fits the training data while test accuracy remains low, and only later begins to generalize. We ask whether this transition can be localized from observed training trajectories before the test accuracy rises, and formulate grokking transition localization as a diagnostic problem with an explicit threshold/FPR/lead-time trade-off. Task-dependent observables are summarized as empirical distributions, mapped to Wasserstein/quantile coordinates, and analyzed by Hankel dynamic mode decomposition (DMD); the resulting reconstruction residual, together with spectrum and effective rank, forms the diagnostic output. On held-out modular-addition Transformer runs, the residual achieves AUROC

\approx

0.93 for grokking-vs-non-grokking discrimination at the run level; under a fixed sustained-threshold operating rule, true-positive alarms can precede onset, with lead time reported jointly with false-alarm rate and uncertainty intervals. Perturbation experiments show that, in the tested

wd=1

pool, high-residual windows exhibit about

3\times

larger short-horizon perturbation deviation than low-residual windows. In a same-data norm-window control, perturbation sensitivity aligns with the residual ordering rather than total-parameter-norm ordering, suggesting that the residual is not merely a total-norm proxy at the window level in the studied

wd=1

dynamics. Norm signals remain strong run-level regime indicators, and log-probability performs best among the observables tested under the current protocol. We position the residual as a window-level monitoring and localization signal in the studied modular-arithmetic Transformer settings, not a universal early-warning predictor or an intervention rule.

Ziyue Wang, Yufeng Ying, Takafumi Kanamori

Institute of Science Tokyo · University of Science and Technology of China

Grokking Transformer Residual Streams Dynamic Mode Decomposition Spectral Transformer Architectures Diagnosis+1

May 6, 2026cs.LG

Shortcut Solutions Learned by Transformers Impair Continual Compositional Reasoning

Identifying and exploiting common features across domains is at the heart of the human ability to make analogies, and is believed to be crucial for the ability to continually learn. To do this successfully, general and flexible computational strategies must be developed. While the extent to which Transformer neural network models can perform compositional reasoning has been the subject of intensive recent investigation, little work has been done to systematically understand how well these models can leverage their representations to learn new, related experiences. To address this gap, we expand the previously developed Learning Equality and Group Operations (LEGO) framework to a continual learning (CL) setting ("continual LEGO"). Using this continual LEGO experimental paradigm, we study the capability of feedforward and recurrent Transformer models to perform CL. We find that BERT, a canonical feedforward Transformer model, learns shortcut solutions that limits its ability to generalize and prevents strong forward transfer to new experiences. In contrast, we find evidence supporting the hypothesis that ALBERT, a recurrent version of BERT, learns a For loop-esque solution, which leads to better CL performance. When applying BERT and ALBERT models to a CL setting that requires composition across experiences, we find that both model families fail. Our investigation suggests that ALBERT models can have their performance drop rescued by use of training strategies that combine data across experiences, but this is not true for BERT models, where a detrimental shortcut solution becomes entrenched with initial training. Our results demonstrate that the recurrent ALBERT model may have an inductive bias better suited for CL and motivate future investigation of the interplay between Transformer architecture and computational solutions that emerge in modern models and tasks.

William T. Redman, Erik C. Johnson, Brian Robinson

Johns Hopkins Applied Physics Lab

Transformer Architectures Replay-Based Continual Learning Continual Learning Bert-Based Models Shortcut Learning

May 6, 2026cs.LG

On Semantic Loss Fine-Tuning Approach for Preventing Model Collapse in Causal Reasoning

Standard fine-tuning of transformer models on causal reasoning tasks leads to catastrophic model collapse, where models learn trivial solutions such as always predicting "Yes" or "No" regardless of input structure. We demonstrate that fine-tuning Gemma 270M on transitivity and d-separation tasks without semantic loss results in 100% collapse rate, with models achieving misleadingly high accuracy (73.9%) while learning no causal reasoning. We propose a semantic loss function with graph-based logical constraints and dynamic lambda scheduling that prevents this collapse. Our approach achieves 70.4% accuracy on transitivity tasks and 68.6% on d-separation tasks with stable, context-dependent predictions, representing a 42.7% improvement over collapsed baselines. Adversarial evaluation on 1,000 structural reasoning samples shows semantic models achieve 67-70% accuracy while collapsed models fail catastrophically at 43-71%. We validate our findings through comprehensive benchmarking on 200,000+ evaluation samples across five model variants, demonstrating that semantic loss is essential and not optional, for stable causal reasoning in transformers.

Pratik Deshmukh, Atirek Gupta

Technical University of Vienna, Austria · HCLTech, Noida, India

Causal Reasoning Collapse Transformer Architectures Ablation Model Fine-Tuning

May 6, 2026cs.CV

Taming Outlier Tokens in Diffusion Transformers

We study outlier tokens in Diffusion Transformers (DiTs) for image generation. Prior work has shown that Vision Transformers (ViTs) can produce a small number of high-norm tokens that attract disproportionate attention while carrying limited local information, but their role in generative models remains underexplored. We show that this phenomenon appears in both the encoder and denoiser of modern Representation Autoencoder (RAE)-DiT pipelines: pretrained ViT encoders can produce outlier representations, and DiTs themselves can develop internal outlier tokens, especially in intermediate layers. Moreover, simply masking high-norm tokens does not improve performance, indicating that the problem is not only caused by a few extreme values, but is more closely related to corrupted local patch semantics. To address this issue, we introduce Dual-Stage Registers (DSR), a register-based intervention for both components: trained registers when available, recursive test-time registers otherwise, and diffusion registers for the denoiser. Across ImageNet and large-scale text-to-image generation, these interventions consistently reduce outlier artifacts and improve generation quality. Our results highlight outlier-token control as an important ingredient in building stronger DiTs.

Xiaoyu Wu, Yifei Wang, Tsu-Jui Fu +3

Rice University · Apple

Diffusion Transformers Visual Tokenizers Vision Transformer Transformer Architectures Image Generation Outliers+1

May 6, 2026cs.LG

Understanding In-Context Learning for Nonlinear Regression with Transformers: Attention as Featurizer

Pre-trained transformers are able to learn from examples provided as part of the prompt without any weight updates, a remarkable ability known as in-context learning (ICL). Despite its demonstrated efficacy across various domains, the theoretical understanding of ICL is still developing. Whereas most existing theory has focused on linear models, we study ICL in the nonlinear regression setting. Through the interaction mechanism in attention, we explicitly construct transformer networks to realize nonlinear features, such as polynomial or spline bases, which span a wide class of functions. Based on this construction, we establish a framework to analyze end-to-end in-context nonlinear regression with the constructed features. Our theory provides finite-sample generalization error bounds in terms of context length and training set size. We numerically validate the theory on synthetic regression tasks.

Alexander Hsu, Zhaiming Shen, Wenjing Liao +1

Department of Mathematics, Purdue University · School of Mathematics, Georgia Institute of Technology

In-Context Learning Transformer Architectures Splines Feature Extraction

May 6, 2026cs.LG

Superposition Is Not Necessary: A Mechanistic Interpretability Analysis of Transformer Representations for Time Series Forecasting

Transformer architectures have been widely adopted for time series forecasting, yet whether the representational mechanisms that make them powerful in NLP actually engage on time series data remains unexplored. The persistent competitiveness of simple linear models such as DLinear has fueled ongoing debate, but no mechanistic explanation for this phenomenon has been offered. We address this gap by applying sparse autoencoders (SAEs), a tool from mechanistic interpretability, to probe the internal representations of PatchTST. We first establish that a single-layer, narrow-dimensional transformer matches the forecasting performance of deeper configurations across commonly used benchmarks. We then train SAEs on the post-GELU intermediate FFN activations with dictionary sizes ranging from 0.5x to 4.0x the native dimensionality. Expanding the dictionary yields negligible downstream performance change (average 0.214%), with large portions of overcomplete dictionaries remaining inactive. Targeted causal interventions on dominant latent features produce minimal forecast perturbation. Across all evaluated settings, we observe no empirical evidence that the analyzed FFN representations rely on strong superposition. Instead, the representations remain sparse, stable under aggressive dictionary expansion, and largely insensitive to latent interventions. These results demonstrate that superposition is not necessary for competitive performance on standard forecasting benchmarks, suggesting they may not demand the rich compositional representations that drive transformer success in language modeling, and helping explain the persistent competitiveness of simple linear models

Alper Yıldırım

Forecasting Backbone Time Series Forecasting Transformer Architectures Mechanistic Interpretability Representational Capacity Language Modeling

May 6, 2026cs.LG

Adaptive Inverted-Index Routing for Granular Mixtures-of-Experts

Mixture-of-experts (MoE) models enable scalable transformer architectures by activating only a subset of experts per token. Recent evidence suggests that performance improves with increasingly granular experts, i.e., many small experts instead of a few large ones. However, this regime substantially increases routing cost, which can dominate computation. We introduce adaptive inverted-index routing for MoE (AIR-MoE), an inverted-index-inspired routing architecture based on vector quantization (VQ). In a first stage, AIR-MoE performs coarse shortlisting by assigning tokens to VQ codewords to construct a candidate set of experts. In a second stage, fine scoring computes exact routing scores restricted to this shortlist. This two-stage procedure approximates true top-k routing while avoiding full expert scoring and, in contrast to prior work, imposing no structural constraints on expert parameters. AIR-MoE serves as a drop-in replacement for standard routers and requires no modifications to the model architecture or loss function. We further provide a lower bound on the mass recall achieved by AIR-MoE that yields insights into its inner workings. Empirically, we demonstrate that AIR-MoE achieves improved performance compared to existing routing approaches in granular MoE settings.

Klaus-Rudolf Kladny, Maximilian Mordig, Bernhard Schölkopf +1

Max Planck Institute for Intelligent Systems, Tübingen · Tübingen AI Center · ETH Zurich +1

Mixture-Of-Experts Large Language Model Routing Transformer Architectures

May 6, 2026cs.CR

On the (In-)Security of the Shuffling Defense in the Transformer Secure Inference

For Transformer models, cryptographically secure inference ensures that the client learns only the final output, while the server learns nothing about the client's input. However, securely computing nonlinear layers remains a major efficiency bottleneck due to the substantial communication rounds and data transmission required. To address this issue, prior works reveal intermediate activations to the client, allowing nonlinear operations to be computed in plaintext. Although this approach significantly improves efficiency, exposing activations enables adversaries to extract model weights. To mitigate this risk, existing works employ a shuffling defense that reveals only randomly permuted activations to the client. In this work, we show that the shuffling defense is not as robust as previously claimed. We propose an attack that aligns differently shuffled activations to a common permutation and subsequently exploits them to extract model weights. Experiments on Pythia-70m and GPT-2 demonstrate that the proposed attack can align shuffled activations with mean squared errors ranging from

10^{-9}

to

10^{-6}

. With a query cost of approximately $1, the adversary can recover model weights with L1-norm differences ranging from

10^{-4}

to

10^{-2}

compared to the oracle weights.

Zhengyi Li, Yakai Wang, Kang Yang +6

Shanghai Jiao Tong University · State Key Laboratory of Cryptology · Shanghai Qizhi Institute

Transformer Architectures LLM Defense Mechanisms Security Bottlenecks

May 6, 2026cs.CL

Anticipating Innovation Using Large Language Models

Forecasting innovation, intended as the emergence of new technological combinations, is a fundamental challenge for science and policy. We show that forthcoming combinations leave an early trace in the collective language of patents, with predictive signals detectable even decades in advance. We show that signal is not attributable to any single inventor, but emerges as a collective shift in how technologies are described across thousands of patents. To this end, we introduce TechToken, a transformer-based model that treats technologies, classified by International Patent Classification codes, as words in its vocabulary, learning the language of technologies by embedding these codes during fine-tuning. We define context similarity between code embeddings as a measure of linguistic convergence and show that it accurately predicts first technological combinations. TechToken also improves general representation quality, outperforming state-of-the-art models across different patent-related tasks.

Enrico Maria Fenoaltea, Filippo Santoro, Giordano De Marzo +2

Institute of Complex Systems (UBICS), Universitat de Barcelona · Centro Ricerche Enrico Fermi (CREF), Via Panisperna 89 A – 00184 Roma · University of Konstanz +1

Technology Large Language Model Forecasting Next-Token Prediction Anticipate Transformer Architectures International Classification Of Disease Coding

May 6, 2026cs.CL

Gyan: An Explainable Neuro-Symbolic Language Model

Transformer based pre-trained large language models have become ubiquitous. There is increasing evidence to suggest that even with large scale pre-training, these models do not capture complete compositional context and certainly not, the full human analogous context. Besides, by the very nature of the architecture, these models hallucinate, are difficult to maintain, are not easily interpretable and require enormous compute resources for training and inference. Here, we describe Gyan, an explainable language model based on a novel non-transformer architecture, without any of these limitations. Gyan achieves SOTA performance on 3 widely cited data sets and superior performance on two proprietary data sets. The novel architecture decouples the language model from knowledge acquisition and representation. The model draws on rhetorical structure theory, semantic role theory and knowledge-based computational linguistics. Gyan's meaning representation structure captures the complete compositional context and attempts to mimic humans by expanding the context to a 'world model'. AI model adoption critically depends on trust and transparency especially in mission critical use cases. Collectively, our results demonstrate that it is possible to create models which are trustable and reliable for mission critical tasks. We believe our work has tremendous potential for guiding the development of transparent and trusted architectures for language models.

Venkat Srinivasan, Vishaal Jatav, Anushka Chandrababu +1

Innospark Ventures & Gyan AI · Gyan AI Inc.

Large Language Model Pretraining Transformer Architectures Neuro-Symbolic Framework

May 6, 2026cs.CC

Average Attention Transformers and Arithmetic Circuits

We analyse the computational power of transformer encoders as sequence-to-sequence functions on vectors. We show that average hard attention can be used to simulate arithmetic circuits if they are given as an input to an encoder. The circuit families that can be simulated this way have constant depth while using unbounded addition, binary multiplication and sign gates. The transformers we use have arithmetic circuits instead of feed-forward networks. With typical average attention the functions they compute are also computed by the same class of circuit families. Our results hold for transformers over the reals, rationals and any ring in between the two.

Lena Ehrmuth, Laura Strieker

Institute for Theoretical Computer Science, Leibniz University Hanover, Hanover, Germany

Transformer Attention Transformer Encoder Transformer Architectures Sequence Modeling

May 6, 2026eess.SP

Learning the Channel Gain from Anywhere to Anywhere via Cross-environment Transformer Estimators

Channel-gain maps provide the channel gain between any two locations in a geographical region. They find numerous applications, from resource allocation and interference control to path planning for autonomous vehicles. Channel-gain map estimation (CGME) is considerably more challenging than conventional radio map estimation (RME) because channel-gain maps are functions over a 6-dimensional input space. This calls for specialized methods, which currently rely on the (inaccurate) radio tomographic model or require a prohibitively large number of measurements since they do not exploit any spatial structure. This paper overcomes this issue by leveraging spatial patterns that channel-gain maps exhibit across environments, as dictated by the laws of physics and typical environmental characteristics (e.g. building materials and layouts). Adopting a metalearning perspective, a transformer-based estimator is proposed to implicitly learn this common structure from measurements collected in multiple environments. This enables CGME in new environments from significantly fewer measurements (five times less in our experiments). To maximize learning efficiency, the transformer is composed with a feature map that enforces the invariances of CGME, such as those following from reciprocity. Numerical experiments corroborate the merits of the proposed estimator relative to existing methods.

Prasenjit Dhara, Daniel Romero

Dept. of Information and Communication Technology, University of Agder, Grimstad, Norway.

Channel Modeling Spatio-Temporal Transformers Transformer Architectures Meta-Learning Cone-Beam Computed Tomography

May 6, 2026cs.LG

FLUID: Continuous-Time Hyperconnected Sparse Transformer for Sink-Free Learning

Continuous-time (CT) Transformers improve irregular and long-range modeling over CT-RNNs by exploiting inputs or outputs embeddings with continuous dynamics. However, the core scaled-dot-product-attention (SDPA) mechanism remains inherently discrete. We propose FLUID (Flexible Unified Information Dynamics), a CT Transformer that incorporates continuous dynamics directly into the attention computation by replacing it with Liquid Attention Network (LAN). LAN reinterprets attention logits as continuous dynamical system and reformulates them as the solution to a linear ODE modulated by input-dependent nonlinear recurrent gates. Theoretically, we establish stability guarantees for LAN dynamics and show that it serves as an interpolating middle ground between SDPA and CT-RNNs, recovering each as special case under well-defined parameterization of its gating functions. LAN also introduces an explicit attention-sink gate to eliminate disproportionate attention mass on uninformative nodes. FLUID replaces standard residual connections with input-dependent Liquid Hyper-Connections to adaptively regulate interlayer information flow. Empirically, we evaluate FLUID on a broad set of learning tasks, including (i) irregular time-series, (ii) long-range modeling, (iii) lane-keeping control of autonomous vehicles, and (iv) learning physical dynamics under a scarce data regime. Across all the tasks, FLUID consistently matches or outperforms CT baselines, achieving improvements of up to 47% in certain scenarios and enhancing generalization under distributional shifts. Additionally, FLUID demonstrates superior noise robustness and a self-correcting inductive bias in autonomous vehicle control. We also provide a detailed analysis of key hyperparameters to guide tuning and show that FLUID occupies an intermediate position among competing approaches in terms of runtime and memory efficiency.

Waleed Razzaq, Yun-Bo Zhao

School of Automation University of Science and Technology China Hefei, Anhui

Transformer Architectures Replay-Based Continual Learning

May 6, 2026cs.LG

Critical Windows of Complexity Control: When Transformers Decide to Reason or Memorize

Recent work has shown that Transformers' compositional generalization is governed by \emph{complexity control}, initialization scale and weight decay, which steers training toward low-complexity reasoning solutions rather than high-complexity memorization. Existing analyses, however, treat complexity control as a single static hyperparameter choice, leaving open \emph{when} during training this control is actually decisive. We show that the memorization-versus-reasoning fate of a Transformer is determined within a sharp, identifiable window of training. On a controlled compositional task we find that (i)~weight decay applied for a single 25%-of-training window matches full-training weight decay in out-of-distribution (OOD) accuracy (

0.93

vs

0.91

); (ii)~holding total regularization budget constant, placing it in the middle of training yields

5{-}9\times

higher OOD accuracy than placing it early; (iii)~the boundary of the critical window is remarkably sharp, window onset shifted by as little as

100

optimization steps causes mean OOD to jump from chance (

0.15

) to reasoning-regime (

0.61

); (iv)~the window's position depends systematically on initialization scale, but the basin of attraction for reasoning solutions \emph{shrinks} at small initialization, contradicting the prevailing recommendation that smaller initialization is uniformly better. We further show that the critical-window phenomenon is task-specific: it does not appear on grokking with modular arithmetic, where properly tuned constant weight decay matches scheduled weight decay.

Sarwan Ali

Columbia University, Irving Medical Center, NY, USA

Transformer Architectures Weight Decay Compositional Generalization Regularization Memorization Reasoning Skills

May 5, 2026cs.AI

The Scaling Properties of Implicit Deductive Reasoning in Transformers

We investigate the scaling properties of implicit deductive reasoning over Horn clauses in depth-bounded Transformers. By systematically decorrelating provability from spurious features and enforcing algorithmic alignment, we find that in sufficiently deep models with a bidirectional prefix mask, implicit reasoning approaches explicit CoT performance across graph topologies and problem widths, though CoT remains necessary for depth extrapolation.

Enrico Vompa, Tanel Tammet

Applied Artificial Intelligence Group Tallinn University of Technology, Estonia

Transformer Architectures Deduction Theorem Proving

May 5, 2026cs.CV

UniCorrn: Unified Correspondence Transformer Across 2D and 3D

Visual correspondence across image-to-image (2D-2D), image-to-point cloud (2D-3D), and point cloud-to-point cloud (3D-3D) geometric matching forms the foundation for numerous 3D vision tasks. Despite sharing a similar problem structure, current methods use task-specific designs with separate models for each modality combination. We present UniCorrn, the first correspondence model with shared weights that unifies geometric matching across all three tasks. Our key insight is that Transformer attention naturally captures cross-modal feature similarity. We propose a dual-stream decoder that maintains separate appearance and positional feature streams. This design enables end-to-end learning through stack-able layers while supporting flexible query-based correspondence estimation across heterogeneous modalities. Our architecture employs modality-specific backbones followed by shared encoder and decoder components, trained jointly on diverse data combining pseudo point clouds from depth maps with real 3D correspondence annotations. UniCorrn achieves competitive performance on 2D-2D matching and surpasses prior state-of-the-art by 8% on 7Scenes (2D-3D) and 10% on 3DLoMatch (3D-3D) in registration recall. Project website: https://neu-vi.github.io/UniCorrn

Prajnan Goswami, Tianye Ding, Feng Liu +1

Northeastern University · Adobe Research

Point Clouds 3D Generation Transformer Architectures Encoders

May 5, 2026cs.CV

RD-ViT: Recurrent-Depth Vision Transformer for Semantic Segmentation with Reduced Data Dependence Extending the Recurrent-Depth Transformer Architecture to Dense Prediction

Vision Transformers (ViTs) achieve state-of-the-art segmentation accuracy but require large training datasets because each layer has unique parameters that must be learned independently. We present RD-ViT, a Recurrent-Depth Vision Transformer that adapts the Recurrent-Depth Transformer (RDT) architecture to dense prediction tasks, supporting both 2D and 3D inputs. RD-ViT replaces the deep stack of unique transformer blocks with a single shared block looped T times, augmented with LTI-stable state injection for guaranteed convergence, Adaptive Computation Time (ACT) for spatial compute allocation, depth-wise LoRA adaptation, and optional Mixture-of-Experts (MoE) feed-forward networks for category-specific specialization. We evaluate on the ACDC cardiac MRI segmentation benchmark in both 2D slice-level and 3D volumetric settings with exclusively real experiments executed in Google Colab. In 2D, RD-ViT outperforms standard ViT at 10% training data (Dice 0.774 vs 0.762) and at full data (0.882 vs 0.872). In 3D, RD-ViT with MoE achieves Dice 0.812 with 3.0M parameters, reaching 99.4% of standard ViT performance (0.817) at 53% of the parameter count. MoE expert utilization analysis reveals that different experts spontaneously specialize for different cardiac structures (RV, MYO, LV) without explicit routing supervision. ACT halting maps show higher compute allocation at cardiac boundaries, and the mean ponder time decreases from 2.6 to 1.4 iterations during training, demonstrating learned computational efficiency. Depth extrapolation enables inference with more loops than training without degradation. All code, notebooks, and results are publicly released.

Renjie He

Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA

Self-Supervised Vision Transformers Dense Prediction Semantic Segmentation Transformer Architectures

May 5, 2026cs.CL

Feature-Augmented Transformers for Robust AI-Text Detection Across Domains and Generators

AI-generated text is nowadays produced at scale across domains and heterogeneous generation pipelines, making robustness to distribution shift a central requirement for supervised binary detectors. We train transformer-based detectors on HC3 PLUS and calibrate a single decision threshold by maximising balanced accuracy on held-out validation; this threshold is then kept fixed for all downstream test distributions, revealing domain- and generator-dependent error asymmetries under shift. We evaluate in-domain on HC3 PLUS, under cross-dataset transfer to the multi-domain, multi-generator M4 benchmark, and on the external AI-Text-Detection-Pile. Although base models achieve near-ceiling in-domain performance (up to 99.5% balanced accuracy), performance under shift is brittle and strongly model-dependent. Feature augmentation via attention-based linguistic feature fusion improves transfer, with our best model (DeBERTa-v3-base+FeatAttn) achieving 85.9% balanced accuracy on M4. Multi-seed experiments confirm high stability. Under the same fixed-threshold protocol, our model outperforms strong zero-shot baselines by up to +7.22 points. Category-level ablations further show that readability and vocabulary features contribute most to robustness under shift. Overall, these results demonstrate that feature augmentation and a modern DeBERTa backbone significantly outperform earlier BERT/RoBERTa models, while the fixed-threshold protocol provides a more realistic and informative assessment of practical detector robustness.

Mohamed Mady, Johannes Reschke, Björn Schuller

Chair of Health Informatics, Technical University of Munich (TUM), Germany · OTH Regensburg, Germany · GLAM – Group on Language, Audio & Music, Imperial College London, United Kingdom

Machine-Generated Text Detection Bert-Based Models Detection Transformer Transformer Architectures Generators

May 5, 2026cs.LG

Transformers with Selective Access to Early Representations

Several recent Transformer architectures expose later layers to representations computed in the earliest layers, motivated by the observation that low-level features can become harder to recover as the residual stream is repeatedly transformed through depth. The cheapest among these methods add static value residuals: learned mixing coefficients that expose the first-layer value projection V_1 uniformly across tokens and heads. More expressive dense or dynamic alternatives recover finer-grained access, but at higher memory cost and lower throughput. The usefulness of V_1 is unlikely to be constant across tokens, heads, and contexts; different positions plausibly require different amounts of access to early lexical or semantic information. We therefore treat early-representation reuse as a retrieval problem rather than a connectivity problem, and introduce Selective Access Transformer (SATFormer), which preserves the first-layer value pathway while controlling access with a context-dependent gate. Across models from 130M to 1.3B parameters, SATFormer consistently improves validation loss and zero-shot accuracy over the static value-residual and Transformer baselines. Its strongest gains appear on retrieval-intensive benchmarks, where it improves over static value residuals by approximately 1.5 average points, while maintaining throughput and memory usage close to the baseline Transformer. Gate analyses suggest sparse, depth-dependent, head-specific, and category-sensitive access patterns, supporting the interpretation that SATFormer learns selective reuse of early representations rather than uniform residual copying. Our code is available at https://github.com/SkyeGunasekaran/SATFormer.

Skye Gunasekaran, Téa Wright, Rui-Jie Zhu +1

UC Santa Cruz · UC Berkeley

Transformer Architectures Early Layers

May 5, 2026cs.LG

Task Vector Geometry Underlies Dual Modes of Task Inference in Transformers

Transformers are effective at inferring the latent task from context via two inference modes: recognizing a task seen during training, and adapting to a novel one. Recent interpretability studies have identified from middle-layer representations task-specific directions, or task vectors, that steer model behavior. However, a lack of rigorous foundations hinders connecting internal representations to external model behavior: existing work fails to explain how task-vector geometry is shaped by the training distribution, and what geometry enables out-of-distribution (OOD) generalization. In this paper, we study these questions in a controlled synthetic setting by training small transformers from scratch on latent-task sequence distributions, which allows a principled mathematical characterization. We show that two inference modes can coexist within a single model. In-distribution behavior is governed by Bayesian task retrieval, implemented internally through convex combinations of learned task vectors. OOD behavior, by contrast, arises through extrapolative task learning, whose representations occupy a subspace nearly orthogonal to the task-vector subspace. Taken together, our results suggest that task-vector geometry, training distributions, and generalization behaviors are closely related.

Hao Yan, Haolin Yang, Yiqiao Zhong

University of Wisconsin–Madison · University of Chicago

Transformer Architectures

May 5, 2026cs.LG

Disentangling Shared and Task-Specific Representations from Multi-Modal Clinical Data

Real-world clinical data is inherently multimodal, providing complementary evidence that mirrors the practical necessity of jointly assessing multiple related outcomes. Although multi-task learning can improve efficiency by sharing information across outcomes, existing approaches often fail to balance shared representation learning with outcome-specific modeling. Hard parameter sharing can trigger negative transfer when task gradients conflict, while flexible sharing may still entangle shared and task-specific signals. To address this, we propose a multi-task framework built on a unified Transformer for multimodal fusion, augmented with Orthogonal Task Decomposition (OrthTD) to split patient representations into shared and task-specific subspaces and impose a geometric orthogonality constraint to reduce redundancy and isolate task-specific signals. We evaluated OrthTD on a real-world cohort of 12,430 surgical patients for predicting four outcomes. OrthTD achieved average AUC (area under the receiver operating characteristic curve) of 87.5% and average AUPRC (area under the precision-recall curve) of 37.2%, consistently outperformed advanced tabular and multi-task methods. Notably, OrthTD achieves substantial gains in AUPRC, indicating superior performance in identifying rare events within imbalanced clinical data. These results suggest that enforcing non-redundant shared and task-specific representations can improve multi-outcome prediction from multimodal clinical data.

He Lyu, Huolin Zeng, Junren Wang +7

Department of Anesthesiology and West China Biomedical Big Data Center West China Hospital, Sichuan University, Chengdu, China · Med-X Center for Informatics, Sichuan University, Chengdu, China · Department of National Key Laboratory of Fundamental Science on Synthetic Vision, Sichuan University, Chengdu, China +4

Multimodal Clinical Data Transformer Architectures

May 5, 2026cs.CL

Benchmarking Logistic Regression, SVM, Naive Bayes, and IndoBERT Fine-Tuning for Sentiment Analysis on Indonesian Product Reviews

The exponential growth of e-commerce platforms in Indonesia has generated a massive volume of user-generated product reviews. Analyzing the sentiment of these reviews is critical for measuring customer satisfaction and identifying product issues at scale. This paper benchmarks traditional Machine Learning (ML) approaches against a Transformer-based Deep Learning model for a three-class sentiment analysis task (positive, neutral, negative) on the Tokopedia Product Reviews 2025 dataset. We implemented Term Frequency-Inverse Document Frequency (TF-IDF) feature extraction coupled with three algorithms: Logistic Regression, Linear Support Vector Machine (SVM), and Multinomial Naive Bayes as robust baselines. Subsequently, we fine-tuned the IndoBERT model (indobenchmark/indobert-base-p1) for contextual sequence classification. To computationally address the severe class imbalance inherent in e-commerce feedback, we applied balanced class weights for the baseline models and engineered a custom weighted cross-entropy loss function within the IndoBERT training loop, following the broader motivation of imbalanced-learning research. Our comprehensive evaluation using Accuracy, Macro F1-score, and Weighted F1-score revealed that the traditional Linear SVC model significantly outperformed the IndoBERT model in our experimental setup, achieving an Accuracy of 97.60% and a Macro F1-score of 0.5510, compared to IndoBERT's 88.70% and 0.5088. Detailed analysis indicates that this performance gap was primarily driven by discrepancies in the data sampling regimes, where baselines utilized the full corpus while the Transformer was constrained to a sampled subset. Finally, we demonstrate the practical viability of our pipeline by deploying the final sentiment classification model as an interactive Gradio web application.

Nabila Zakiyah Zahra, Salwa Farhanatussaidah, Nasywa Nur Afifah +3

Department of Data Science Institut Teknologi Sumatera (ITERA) · Department of Informatics Institut Teknologi Sumatera (ITERA)

Term Frequency-Inverse Document Frequency Indonesian E-Commerce Reviews Sentiment Imbalanced Classification Logistic Regression Model Fine-Tuning+1