Parameter Sharing

Most long-horizon forecasters emit the complete future in one pass. We ask whether a compact operator can instead be reused to emit contiguous future blocks, trading stored parameters for sequential computation. Recursive Functional Decomposition (RFD) factorizes the horizon mapping into three parameter-shared block predictions. The blocks occupy disjoint future intervals and are concatenated, not added as trend, seasonal, or residual components, and each emitted block joins the rolling context that conditions the next application. In controlled three-seed experiments, removing forward feedback hurts all four evaluated datasets while stopping gradients through it changes little, so the useful signal is the evolving forecast state itself; reapplying the shared operator helps consistently, but by less than widening it. The model uses 0.08--0.46M parameters, yet its sequential channel-wise execution is slower and more memory-intensive than an iterative baseline at high channel count. In a frozen long-context evaluation against ten strong baseline families, RFD records lower test error in 23 of 24 cells; the comparison is cross-protocol, with RFD using 512/720-step context and the baselines using 96.

Personalized large language models often require a complete adaptation state for each user. However, this paradigm scales poorly as the user population grows. We revisit this design through the lens of personalization capacity allocation: how much adaptation capacity can be shared across users, how the shared capacity should be composed, and how much must remain user-specific. We answer them through three complementary empirical analyses. We find that independent user adapters contain substantial cross-user reusable structure, that the utility of reusable directions reflects both user relevance and variation across queries, and that user histories provide transferable signals for compact individual correction. Motivated by these findings, we propose LINEUP. It learns a bank of reusable low-rank personalization factors, composes them through user-conditioned recall and query-dependent calibration, and restricts target-user adaptation to a tiny user code over a shared correction space. This design decouples expressive personalization capacity from per-user trainable state. Each target user optimizes only eight scalars, while all shared components remain fixed. By comparison, the evaluated private-LoRA configuration uses 4.19 million per-user parameters. Our theoretical analysis gives a finite-step, finite-history risk bound and sufficient conditions for user-code refinement to improve on history initialization. Across six tasks spanning personalized classification, prediction, and generation, LINEUP leads on all 12 metrics, each averaged over three independent runs (e.g., reducing LaMP-3 RMSE by 11.4% relative to the strongest baseline). It maintains advantages under limited history. These results show that rich personalization can be supported primarily by reusable, conditionally composed shared capacity, while independent user adaptation remains confined to a tiny correction state.

Memory constraints in artificial intelligence motivate accurate function approximation with fewer parameters. We study looping, which repeatedly composes one update function with shared parameters; each output becomes the next input. A looped Transformer, for example, reuses one block, whereas its conventional untied counterpart uses separately parameterized blocks. We compare their parameter requirements for a given worst-case approximation accuracy, or equivalently, their approximation accuracy under a common budget limiting distinct trainable coefficients. We then ask whether this representational parsimony improves statistical accuracy. For general likelihood models, we establish an upper squared Hellinger risk bound for looped sieve maximum likelihood and a minimax lower bound for the jointly tuned untied family. Further loop iterations improve the approximation bound without adding parameters, while increasing computation and the fitted-class complexity bound. For targets of known Hölder smoothness, looped residual feedforward networks and post-layer-normalized Transformers attain the minimax polynomial rate up to logarithmic factors with a fixed number of bounded real parameters. At sufficiently large fixed budgets, looped worst-case risk vanishes while optimal worst-case untied risk remains bounded away from zero. The loop-to-untied risk ratio also tends to zero under specified growing-budget conditions. Regression, binary response, and energy-based generative models illustrate the theory.

While Parameter-Efficient Fine-Tuning (PEFT) has substantially reduced the hardware cost of adapting Large Language Models (LLMs) by decreasing the number of trainable parameters, recent studies have sought to further improve PEFT through parameter sharing. However, these approaches either employ uniform parameter sharing across layers, which can delay convergence, or rely on dynamic masking strategies, which add computational overhead. The potential of sharing patterns inspired by the inherent hierarchical structure of Transformer architectures remains unexplored in PEFT. To address this gap, we introduce SAPE (Sandwich Adapters for Parameter Efficiency), a PEFT framework based on a sandwich-style hard weight-sharing topology. SAPE routes intermediate Transformer layers through balanced shared group adapters while strictly isolating the input embedding and final projection boundary transformations to prevent gradient interference. This design significantly reduces memory consumption while eliminating the computational overhead associated with dynamic parameter-sharing methods. Extensive evaluations across encoder-only and causal decoder architectures demonstrate that SAPE achieves state-of-the-art performance in low-parameter regimes. On natural language understanding, SAPE outperforms proPETL on RoBERTa-large while utilizing only 10% of the baseline's parameter budget. On natural language generation and world knowledge reasoning with LLaMA-3.2 (3B) under a strict ~0.6M parameter constraint, SAPE outperforms AdaLoRA, yielding absolute improvements of +4.85% on GSM8K and +3.11% on CommonsenseQA. Furthermore, through comprehensive topological ablations, we formalize an inherent capacity trade-off: while hard parameter sharing strongly regularizes semantic generalization, it slightly smooths the sharp layer-wise transformations required for rigid multi-step arithmetic reasoning.

Federated learning (FL) must serve devices with varying computational capabilities. A fixed model cannot suit all devices, while training one model per deployment limit is costly. Federated supernet training instead learns one elastic model with differently sized subnetworks, then deploys a suitable one to each device. When client inference budgets differ, however, parameters exclusive to high-cost subnetworks are reachable by fewer clients. We propose FEAST, a federated shared-space training framework that counters this imbalance by jointly training multiple subnetworks within each client's limit. Budget-tailored sub-supernet routing sends only the relevant supernet portion, and sparse aggregation merges the returned parameter slices. The trained supernet directly serves the subnetworks used during federation and supports post-hoc extraction of additional subnetworks without federated retraining. We further show that independently assigning clients' training-data volumes and inference budgets can distort accuracy--inference-cost comparisons in heterogeneous FL simulations, and introduce a one-parameter $γ$-allocation protocol to control this coupling. In our experimental setup, the SuperFedNAS and DeepFedNAS supernet training procedures remain near chance at 25M and reach at most $17.09\%$ at $596$M inference MACs; FEAST reaches $71.06\%$ at $596$M, $2.4$ points above the strongest model-heterogeneous weight-sharing baseline at its largest tier. Across CIFAR-100, CINIC-10, and TinyImageNet-200, FEAST achieves the highest population-averaged accuracy among the evaluated weight-sharing methods when each client receives its largest affordable subnetwork. Sub-supernet routing reduces aggregate model-parameter traffic by $6.8\times$ relative to full-supernet transmission.

Existing scaling strategies for Multimodal Large Language Models (MLLMs) typically expand either model parameters or sequential inference computation, incurring substantial memory or latency overhead. More importantly, most existing methods fail to alter the rigid, fixed computation allocation between the Vision Transformer and the Large Language Model components, limiting task-specific optimization. To address this, we introduce the Parallel Vision-Language (ParVL) scaling framework for MLLMs, which scales parallel computation by reusing the existing ViT and LLM backbone parameters across multiple vision and language branches. This framework raises a central question: given a fixed backbone parameter budget, how should additional shared-backbone computation be allocated between the vision and language modalities? We instantiate each parallel computational stream with branch-specific prefix parameters over a shared backbone, and train the entire model end-to-end via full-parameter supervised fine-tuning on roughly 13B tokens. We systematically study the computation-allocation trade-off between the ViT encoder and LLM decoder. ParVL improves overall multimodal performance over same-recipe single-branch baselines, and the best evaluated vision--language allocation varies across tasks. Code is available at https://github.com/YangYangGirl/ParVL.

Designing effective feature extractors is critical for blind signal analysis tasks such as automatic modulation recognition (AMR), signal scheme recognition (SSR), and \color{black} signal structure parsing (SSP). In this work, we propose dual-channel neural network (DualNN) that efficiently exploits complex-valued signals through parameter sharing across IQ channels. Unlike traditional real-valued or complex-valued models, DualNN is a groundbreaking framework which shares the network parameters for processing the real and imaginary parts of the complex-valued signals, and is theoretically shown to reduce generalization error while preserving expressive capacity. Specifically, we propose a novel Transformer-based architecture to implement DualNN, called Dualformer. The Dualformer segments input signals into patch-level tokens and captures multi-granularity features, enabling robust performance across diverse signal analysis tasks. Furthermore, we conduct extensive experiments comparing Dualformer with three Transformer-based baselines and four conventional DL-based approaches. Results demonstrate consistent performance improvements on AMR, SSR, and SSP tasks. Besides, the modular design of DualNN allows it to generalize well to blind signal processing tasks such as blind source separation and low-SNR spectrum sensing. This work paves the way for a broader application of DualNN architectures in unsupervised and weakly supervised complex-valued signal analysis scenarios.

Personalized federated learning (PFL) is one of the important approaches in federated learning for addressing statistical heterogeneity while enabling client-specific adaptation. Many PFL methods split the model into shared and personalized parameters, which are jointly trained on each client. However, this creates an optimization issue: shared parameters are updated by clients optimizing different local objectives, which can lead to inconsistent shared updates and weaken the shared representation. To address this problem, we propose Federated Shared Parameter Correction (FedSPC), a modular correction method for PFL. FedSPC applies control-variate correction only to the shared parameters of a given PFL method, while leaving personalized parameters unchanged. It can be integrated into three common PFL settings: shared feature extractors, shared classifiers, and fully shared models with local regularization. Experiments on CIFAR-100 and Tiny-ImageNet with ViT, ResNet-34, and VGG-11 show that FedSPC improves performance across representative PFL methods, including FedPer, FedRep, FedBABU, LG-FedAvg, and Ditto.

Softmax attention is the cornerstone of modern large language models, but its memory scales linearly and compute quadratically with sequence length. Linear recurrent models, such as linear attention and state space models, have become widely studied as alternatives to attention due to their linear compute and constant memory. While these sub-quadratic token mixing methods, or mixers, achieve promising efficiency gains and competitive results on a wide range of benchmarks, current linear recurrent models still lag behind on tasks that require long-context retrieval or in-context learning. A growing body of work studies hybrid architectures that attempt to mitigate these trade-offs by statically interleaving or merging attention and recurrent blocks. In this work, we explore a new axis of developing hybrid models: across the token sequence. We propose Oryx, a hybrid model that can, throughout a sequence, flexibly switch between different mixers, for example quadratic attention for rich context utilization and linear recurrences for efficient generation. Oryx ties at least 90% of its parameters across mixers, enabling attention and recurrent modes to operate over shared internal representations. We validate our design with Mamba-2 and Gated DeltaNet variants, up to 1.4B models. Under fixed token budgets and a mixed-training strategy, Oryx achieves comparable or better performance than its single-mixer baselines. At the 1.4B scale, all instances of Oryx outperform their respective baselines by at least 0.7 percentage points on averaged language modeling tasks. On retrieval tasks, Oryx achieves performance comparable to the Transformer baseline even when processing only a tiny fraction (<10%) of the tokens in attention mode. These results suggest that attention and linear recurrent models can share internal representations, and motivate sequence-axis hybridization as a promising direction.

State Space Models (SSMs) are increasingly deployed in the Edge because they offer, at comparable performance, a smaller memory/training/inference footprint, compared to Large Language Models (LLMs). These three advantages are a direct consequence of the time recurrence inherent in the SSMs architecture. Here, we further improve this recurrent architecture by positively answering two previously underexplored, orthogonal questions: (1) Can we reduce SSMs memory-footprint without any performance penalty, by also employing depth recurrence? (2) Can we increase SSMs performance by using a fixed and consistent time-granularity across all tasks? The first question is somewhat unexpected, given that SSMs are already recurrent. However, the orthogonal depth recurrence further decreases SSMs memory footprint. We show that a looped SSM with $k$ parameters adaptively iterated $M$ times, achieves a performance comparable to a standard SSM with $k \cdot L$ independent parameters, where $M \leq L$. The second question is also unexpected given the time-recurrent nature of the SSMs architecture. However, it makes perfect sense for the time-parallel training of SSMs on the entire input sequence. We show that concatenating time steps for lower-dimensional sequence elements, or flattening and re-chunking the joint feature-time dimension for high-dimensional ones, can improve the baseline by enhancing the way information is presented to the model. Our results for both extensions lead to consistent benefits across four representative SSM architectures: LRU, S5, LinOSS, LrcSSM.

Synthetic Aperture Radar (SAR) image recognition is vital for disaster monitoring, military reconnaissance, and ocean observation. However, large SAR image sizes hinder deep learning deployment on resource-constrained edge devices, and existing lightweight models struggle to balance high-precision feature extraction with low computational requirements. The emerging Kolmogorov-Arnold Network (KAN) enhances fitting by replacing fixed activations with learnable ones, reducing parameters and computation. Inspired by KAN, we propose Light-ResKAN to achieve a better balance between precision and efficiency. First, Light-ResKAN modifies ResNet by replacing convolutions with KAN convolutions, enabling adaptive feature extraction for SAR images. Second, we use Gram Polynomials as activations, which are well-suited for SAR data to capture complex non-linear relationships. Third, we employ a parameter-sharing strategy: each kernel shares parameters per channel, preserving unique features while reducing parameters and FLOPs. Our model achieves 99.09%, 93.01%, and 97.26% accuracy on MSTAR, FUSAR-Ship, and SAR-ACD datasets, respectively. Experiments on MSTAR resized to $1024 \times 1024$ show that compared to VGG16, our model reduces FLOPs by $82.90 \times$ and parameters by $163.78 \times$. This work establishes an efficient solution for edge SAR image recognition.

In low-resource multilingual speech-to-text translation, uniform architectural sharing across languages frequently introduces representation conflicts that impede convergence. This work proposes a principled methodology to automatically determine layer-specific sharing patterns by mining training gradient information. Our approach employs three distinct analysis strategies: distance-based language clustering, self/cross-task divergence metrics for capacity allocation, and joint factorization coupled with canonical correlation analysis for subspace alignment. Extensive evaluation across four language pairs (using the SeamlessM4T-Medium architecture) demonstrates persistent improvements in translation quality metrics.

Implicit Neural Representations (INRs) have emerged as a promising paradigm for video representation and compression. However, existing multi-scale INR generators often suffer from significant parameter redundancy by stacking independent processing blocks for each scale. Inspired by the principle of scale self-similarity in the generation process, we propose SRNeRV, a novel scale-wise recursive framework that replaces this stacked design with a parameter-efficient shared architecture. The core of our approach is a hybrid sharing scheme derived from decoupling the processing block into a scale-specific spatial mixing module and a scale-invariant channel mixing module. We recursively apply the same shared channel mixing module, which contains the majority of the parameters, across all scales, significantly reducing the model size while preserving the crucial capacity to learn scale-specific spatial patterns. Extensive experiments demonstrate that SRNeRV achieves a significant rate-distortion performance boost, especially in INR-friendly scenarios, validating that our sharing scheme successfully amplifies the core strengths of the INR paradigm.