Parameter-Efficient Fine-Tuning

Also known as PEFT

Ask about this topic

Overview

Parameter-efficient fine-tuning studies how to adapt pretrained models while training only a small subset of weights or adding lightweight modules, reducing resource costs and sometimes helping preserve existing capabilities. Common approaches include low-rank updates such as LoRA and quantized QLoRA, adapters and soft prompts, selective layer or rank tuning, and sparse or routed experts. Methods are evaluated on language benchmarks such as GLUE and mathematical reasoning, as well as vision, speech, multimodal, and continual-learning tasks, often with limited data or hardware. A central tension is that fewer trainable parameters do not necessarily mean lower memory use or faster training, and accuracy, robustness, and retention can vary with the model, task, and placement of updates.

Latest papers

CardsList

610 papers75 papers in the last four weeks, up 12% on the four weeks before. 0.4% of all new papers.

Jun 11, 2026

EPM-JEPA: Operator-Side Experience Modulation in JEPA-Family World Models

JEPA-family world models use a static predictor whose weights do not adapt when test-time dynamics diverge from training. We compare two mechanisms for incorporating accumulated experience into a JEPA predictor under distribution shift: operand-side injection, where a compressed experience representation is added as a residual to the predictor's hidden state (EI-JEPA), and operator-side modulation, where the same representation generates low-rank weight deltas via LoRA applied to the predictor's weights (EPM-JEPA). On a pre-registered comparison (Moving MNIST, gravity shift), EPM-JEPA (D_shift^{n=50} = 0.7848 +/- 0.0078, three seeds) differs from EI-JEPA (0.8238) by delta = 4.74% - Outcome C: a null result - by our stated criterion, a valid outcome. As a secondary, non-pre-registered observation, EPM-JEPA improves 1.90% over a no-memory baseline (0.8000), consistently across seeds, while EI-JEPA underperforms the baseline, indicating the benefit is specific to weight-level modulation. Our primary contribution is a mechanism analysis: the D_shift^{n=50} trajectory reflects three independent dynamical processes - buffer cycling, EMA target drift, and an intrinsic LoRA settling transient of +0.021 - rather than convergence to equilibrium. These findings motivate PEM-JEPA, a physics-grounded successor addressing this dynamical-peak limitation.

Vedant Pandya

School of Artificial Intelligence and Data Engineering (SAIDE) Indian Institute of Technology Jodhpur

Test-Time AdaptationWorld Model LearningJoint-Embedding Predictive ArchitectureTemporal Distribution ShiftParameter-Efficient Fine-Tuning
AskPDF
Jun 11, 2026

LoRA-Muon: Spectral Steepest Descent on the Low-Rank Manifold

Low-Rank Adaptation (LoRA) significantly reduces compute and memory costs for finetuning Deep Learning models but is often harder to tune than dense training: when using factor-wise optimizers such as AdamW, it is sensitive to initialization choices, its optimal learning rates transfer poorly across ranks, and it often fails to beat dense baselines. We derive LoRA-Muon by applying the Muon optimizer's spectral steepest-descent rule to the low-rank setting. Along with our split weight-decay rule, our main claim is that LoRA-Muon is a good low-rank proxy for full-rank Muon and Shampoo-family optimizers. Its optimal learning rates transfer across rank, width, depth, and factor-rescaling. In our compute-matched TinyShakespeare study, a rank-22 proxy recovers the dense best tested learning rate, and a rank-3232 LoRA-Muon run attains lower mean validation loss than the dense baseline in the seed-averaged sweep. We further show that the Spectron optimizer depends on arbitrary factor scaling, so it would likely be a poor fit when finetuning starts from badly imbalanced factors, and that LoRA-RITE's simplified QR-coordinate core implements the same spectral update. LoRA-Muon computes that update without QR-decomposition and avoids storing second moments, making it more accelerator-friendly and memory-efficient.

Franz Louis Cesista, Katherine Crowson, Cédric Simal +1

Ateneo de Manila University · EleutherAI · NaXys +1

Deep Learning OptimizationRiemannian OptimizationMuon OptimizerLow-Rank AdaptationParameter-Efficient Fine-Tuning
AskPDF
Jun 10, 2026

Beyond Prompt and Parser Engineering: Soft-Prompt Tuning for Automated LLM Evaluation

Benchmark scores often misrepresent a large language model's (LLM's) knowledge, because they rely, e.g., on the model's ability to follow specific formatting requirements. This especially penalizes base models that may know the correct answers but lack the ability -- typically introduced in post-training -- to structure them as instructed. To overcome this, we propose an efficient, automated, and architecture-agnostic method for model evaluation based on soft-prompt tuning. By optimizing only 10 soft-prompt vectors (~0.0006% parameters for a 7B model) over a short tuning period, we adapt models to specific benchmark formats, closing gaps in format-following and ensuring that underlying knowledge is accurately reflected in benchmark scores. We evaluate soft-prompt tuning across 7 models and 11 datasets. Our results show that (a) soft-prompt tuning saturates format-following within 150 steps (~1200 samples) while preserving knowledge accuracy, (b) it closely matches strong evaluation-harness baselines on well-studied benchmarks while requiring no manual benchmark-specific prompt or parser design, (c) it improves evaluation signal in settings where baselines are harder to transfer, including long-context datasets and instruction-tuned variants with different output styles, and (d) it produces cleaner outputs that improve agreement between exact match scoring and independent LLM judges. Overall, we propose a lightweight, automated, and flexible protocol based on limited-budget soft-prompt tuning for evaluating LLM knowledge.

Selen Erkan, Bastian Boll, Kristian Kersting +2

Aleph Alpha Research Lab1141 · TU Darmstadt Hessian.AI Lab1141

LLM EvaluationInstruction FollowingParameter-Efficient Fine-TuningTest-Time Prompt Tuning
AskPDF
Jun 10, 2026

Fine-tuning Multi-modal LLMs with ART: Art-based Reinforcement Training

Adapting Large Language Models (LLMs) to specific tasks usually involves fine-tuning: Low-Rank Adaptation (LoRA) edits the weights, while soft prompting prepends trainable virtual tokens. Both change how an already-optimized model is served, so serving many task-specific adapters at high throughput requires adapter-aware infrastructure. We propose Art-Based Reinforcement Training (ART), which optimizes a single auxiliary input image to specialize a frozen Multimodal Large Language Model (MLLM) for a specific task. At inference time, ART simply includes this image in each standard multimodal prompt, requiring no changes to the serving engine. We compare ART with LoRA and soft prompting across five reasoning benchmarks for four small MLLMs from three architectures. While different methods perform best for different tasks, ART ranks first or second in 10 of 20 benchmark-MLLM pairs, with its best results on the two math word problem benchmarks (GSM8k and SVAMP). It is significantly ahead of the next best adapter in 2 cases, within the noise of the next best adapter in 6, and significantly above the text-only baseline in 18. ART therefore reaches accuracy close to LoRA and soft prompts while serving faster than LoRA. In the vLLM serving stack, ART reduces throughput by 6% to 14% relative to text-only requests, compared to LoRA's 13% to 93% slowdown. Beyond efficient serving, ART artifacts are plain images, making them easy to store, distribute, and share. This approach keep task-specific adaptation outside the model weights, so the same frozen model can support different tasks without separate model variants.

Michal Chudoba, Sergey Alyaev, Petra Galuscakova +1

University of Stavanger, Stavanger, Norway · NORCE Research, Bergen, Norway

LLM Fine-TuningMultimodal Large Language ModelsMultimodal Foundation ModelsVisual Prompt LearningParameter-Efficient Fine-Tuning
AskPDF
Jun 10, 2026

Substrate Asymmetry in User-Side Memory: A Diagnostic Framework

User-side memory in LLMs is typically scored as a single "personalization" capability: given a user's history, is the output more user-aware? We show this aggregate metric hides opposite-direction failures. Memory factorises into at least three orthogonal axes -- behavioral consistency (style, voice), factual presence (recall facts in history), and factual absence (abstain when a fact is absent) -- and no single substrate wins all three. Comparing per-user gamma-LoRA (a small LoRA adapter trained on each user's history; gamma denotes per-user, not per-task) against BGE-large dense top-K retrieval on a controlled 50-user synthetic corpus and a real-data probe (LaMP-3), we find gamma-LoRA decisively wins behavioral style while RAG decisively wins factual absence -- and the same query-projection cells in attention layers 21-35 causally load-bear both effects in opposite directions (zeroing those LoRA weights raises absence-probe TPR by +33 pp and drops presence-probe TPR by 20 pp). On the more heavily RLHF-tuned Llama-3.1-8B-Instruct the asymmetry strengthens, not heals: parametric memory's behavioral advantage collapses while its absence-calibration deficit against retrieval widens -- an alignment tax on parametric user-memory. On real-data LaMP-3, gamma-LoRA underperforms a majority baseline; a 9-condition mitigation sweep diagnoses this as instruction-following collapse, not substrate failure (a 9x2 cross-product shows the eval-time {1..5} logit mask drives main_acc to >=0.995 on every recipe), and the best training-time fix replicates bit-identically on Llama. Finally, substrate-selection routing is question-classification, not calibration: a 110M DistilBERT on the question text alone beats every logit-based router. We contribute the diagnostic framework, the diagnosed real-data negative, the alignment-tax replication, and the routing-as-classification finding.

Youwang Deng

EpistemicaLab — Independent Research

Retrieval-Augmented GenerationLLM PersonalizationPersistent Memory for Language ModelsLLM MemoryMemorization in Generative ModelsLLM Abstention+1
AskPDF
Jun 10, 2026

Parameter-Efficient Adapter Tuning for Tabular-Image Multimodal Learning

Tabular-image multimodal learning aims to improve predictive modeling by jointly using structured tabular attributes and visual data. Although pretrained encoders provide strong modality-specific representations, full fine-tuning can be computationally expensive, while keeping encoders frozen may limit task-specific adaptation. We propose the Tabular-Image Adapter (TI-Adapter), a modality-specific adapter-based fine-tuning framework for efficient multimodal adaptation. TI-Adapter freezes the pretrained tabular encoder and learns an adapter after the extracted tabular embedding, while adapting the image branch with embedding-level and bottleneck-level adapters instead of full fine-tuning. Experiments on 20 tabular-image datasets show that TI-Adapter achieves competitive or better predictive performance than full fine-tuning while using substantially fewer trainable parameters. Ablation studies further demonstrate the importance of adapter placement for balancing performance and practical efficiency.

Jiaqi Luo

School of Mathematical Sciences, Soochow University, No.1 Shizi Street, Suzhou, 215006, Jiangsu Province, China

Fine-TuningAdapter TuningTabular MLMultimodal LearningParameter-Efficient Fine-Tuning
AskPDF
Jun 9, 2026

GRAFT: Gain-Recalibrated Adapters for Transformer-Based Neural Population Activity Modeling

Neural population activity models can recover rich temporal structure from binned spikes, but their read-in and readout layers often remain tied to a fixed set of recorded neurons. This coupling limits reuse in long-term brain-computer interfaces, where recorded neuron identities, counts, and response statistics can change across days. We introduce GRAFT, a Transformer-based neural population activity model that separates reusable temporal dynamics from a recalibratable neuron interface. The neuron interface controls how recorded neurons enter and leave the shared backbone, and auxiliary gain and positional mechanisms support neural activity modeling inside the Transformer. On MC Maze under the standard NLB'21 protocol, GRAFT reaches 0.3866 co-bps as an ensemble, setting a new state of the art on the primary co-bps metric among public and reported NLB'21 results. In a cross-day protocol constructed from the NLB'21 MC Maze dataset series, GRAFT recalibrates from MC Maze to the scaled MC Maze datasets (Large/Medium/Small) by updating only 9.21% of parameters, reaching 0.3749, 0.3112, and 0.3152 co-bps with restricted target-day support sets. These results show that the same interface-backbone separation supports both strong Transformer-based neural population activity modeling and data-efficient cross-day recalibration.

Xiangsheng Ge, Yang Xie

Shanghai Jiao Tong University

Brain-Computer InterfacesTransformerNeural EncodingParameter-Efficient Fine-Tuning
AskPDF
Jun 9, 2026

Small Data, Big Noise: Adversarial Training for Robust Parameter-Efficient Fine-Tuning

Parameter-Efficient Fine-Tuning (PEFT) has become essential for adapting foundation models to downstream NLP tasks. However, current PEFT methods often struggle with robustness to noise and performance degradation on limited training data. We propose SDBN (Small Data Big Noise), a unified framework that brings adversarial training to PEFT - a combination that remains less studied in the PEFT setting despite its complementary strengths - to enhance model robustness and generalization, outperforming alternative approaches. We also introduce two variants of the method that use discrete uncertainty sets: SDBN-h, which enumerates character-level edits and selects worst-case variants using gradients, and SDBN-p, which uses LLM-generated variants for robust optimization in generative tasks. Experiments across multiple benchmarks reveal substantial improvements, particularly in low-resource settings and under both word-level and character-level corruptions. This framework addresses the less explored intersection of adversarial training and parameter-efficient adaptation, without introducing additional parameters or only modest computational overhead, making PEFT deployments more reliable in real-world scenarios where data scarcity and linguistic variability often coexist

Eitan Cohen, Idan Simai, Uri Shaham

Bar-Ilan University · Ramat Gan, Israel

Adversarial TrainingNeural Network RobustnessAdversarial Robustness of Language ModelsNoise RobustnessParameter-Efficient Fine-Tuning
AskPDF
Jun 9, 2026

5% > 100%: Flatness Preference is All You Need for Multimodal Parameter-Efficient Fine-Tuning

Parameter-Efficient Fine-Tuning (PEFT) methods provide a streamlined and efficient tool for adapting large models to domain-specific multimodal downstream tasks. Although these methods proved their tangible effects in practice, their principal aspects remain under-explored. Therefore we remain curious about the underlying generalization mechanisms in various PEFT methods and how they can be further enhanced. In this paper, we reveal the flatness preference widely present in various PEFTs, where a small fraction of sharp dimensions dominates the generalization of PEFT. This finding suggests an appealing possibility: we may be satisfied with a better generalization by merely attending to this small fraction of sharp dimensions instead of all of them. Furthermore, we propose Flatness Preference Optimization (FlatPO) to flatten these key sharpness dimensions, leading various PEFTs toward better generalization. Extensive experiments demonstrate the effectiveness of our findings and the proposed method. Code is available at https://github.com/Can-Lin/FlatPO.

Yifan Zhu, Can Lin, Hangjie Yuan +4

School of Computer Sciences, Beijing University of Posts and Telecommunications, Beijing 100876, China · College of Computer Science and Technology, Zhejiang University, Zhejiang 310027, China · ETH Zürich, Switzerland

Deep Learning OptimizationFlat MinimaParameter-Efficient Fine-Tuning
AskPDF
Jun 9, 2026

Advancing the State-of-the-Art in Empirical Privacy Auditing

Parameter-efficient fine-tuning of large language models (LLMs) can exhibit problematic memorization of individual training examples. Empirical privacy auditing (EPA) quantifies this risk by measuring realistic data leakage on membership inference (MI) or reconstruction attacks. A key challenge in EPA is designing ``canary'' examples that are mixed with the privacy-sensitive training data. We propose generating synthetic canaries via high-temperature sampling (T≥0.8T \geq 0.8) from LLMs, using prompts tailored to the privacy-sensitive training data. These canaries act as high-influence outliers, ensuring high identifiability and hence strong audits. Further, since the canaries are themselves non-private, they are inspectable and can be inserted with repetition without jeopardizing the privacy of the real data. An important use of models fine-tuned on privacy-sensitive data is the generation of synthetic data. This also comes with privacy risk. We introduce a powerful synthetic data audit based on fine-tuning an auxiliary model on the synthetic data. Auditing the auxiliary model for the original canaries then provides a strong estimate of the privacy leakage through the synthetic data. Finally, leveraging our strong auditing methodologies, we perform a systematic investigation into the interacting effects of model capacity and canary entropy on memorization.

Nicole Mitchell, Galen Andrew, Arun Ganesh +2

Google Research San Francisco, CA · Google Research Seattle, WA

Data LeakagePrivacy AuditingLLM AuditingPrivacy Leakage in Language ModelsParameter-Efficient Fine-Tuning
AskPDF
Jun 9, 2026

A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models

Though wireless foundation models (WFMs) have shown strong potential in learning universal channel representations, their adaptation to various downstream tasks remains constrained by existing paradigms. Fine-tuning strategies introduces substantial computational and storage overhead, while frozen feature extraction leads to sub-optimal performance across diverse downstream tasks. To address this issue, we propose a unified adaptive feature composition framework for multitask generalization in WFMs, where the key component is the Routing Adapter for Feature Composition (RAFC). Instead of extracting only the final-layer output, this router treats the hidden states from different Transformer depths as a reusable pool of multi-level hidden features, and employs a lightweight task-driven feature composition network to generate layer-wise aggregation weights, then adaptively combine hierarchical representations through weighted summation. This design enables each downstream task to access suitable mixture of low-, mid-, and high-level wireless features without modifying the pretrained backbone. Extensive experiments on four representative wireless tasks demonstrate that RAFC consistently outperforms conventional adaptation baselines while introducing fewer than 50K additional parameters. Moreover, the learned routing weights provide interpretable evidence of task-specific layer preferences, making the proposed framework a low-complexity, scalable, and explainable interface for adapting WFMs to diverse downstream scenarios.

Yuxuan Shi, Tingting Yang, Kangning Ma +4

Department of Broadband Communication, Pengcheng Laboratory, Shenzhen, 518000, China

Wireless Foundation ModelsAdaptive Model RoutingMulti-Task LearningParameter-Efficient Fine-Tuning
AskPDF
Jun 8, 2026

Fisher-Guided Progressive Parameter Selection for Adaptive Fine-Tuning

Parameter-efficient fine-tuning often selects trainable parameters before adaptation using architectural heuristics, without accounting for their varying importance during training. We introduce FisherAdapTune, which progressively selects parameter groups based on temporal drift in their Fisher information. Under a local Gaussian approximation, we bound the divergence between the fine-tuned posterior and pretrained prior by accumulated Fisher-weighted update costs, motivating curvature-aware selection. FisherAdapTune measures Jensen-Shannon distance between successive Fisher-value distributions and uses an adaptive threshold to freeze stabilized groups. Across VTAB-1k classification tasks, it achieves the highest macro Top-1 accuracy among the compared methods with a smaller average trainable set than full fine-tuning. Across four segmentation backbones, it maintains competitive in-distribution performance and improves zero-shot transfer in several settings. Its selections reveal architecture-dependent patterns where input, output, and normalization parameters can remain trainable as attention and MLP groups freeze at different rates. The results support Fisher structural drift as a task-dependent signal for allocating updates during adaptation. We release our code publicly to enable further application of our proposed approach.

Ghodsiyeh Rostami, Po-Han Chen, Mahdi S. Hosseini

Concordia University · Concordia University Mila - Québec AI Institute

Fine-TuningVision Foundation Model AdaptationSelective Fine-TuningFisher InformationParameter-Efficient Fine-Tuning
AskPDF
Jun 8, 2026

LiteVSR: Lightweight Adaptation of Frozen Diffusion Transformers for Video Super-Resolution

Adapting large-scale pre-trained video generators for Video Super-Resolution (VSR) in novel domains remains computationally prohibitive. Methods that reformulate generation as direct Low-Quality to High-Quality mappings deviate from the original generative formulation, demanding extensive fine-tuning. ControlNet-style adapters lose their efficiency under modern Diffusion Transformers since the absence of encoder-decoder hierarchy forces duplication of the entire backbone. We observe that flow matching offers a principled alternative for cross-domain VSR adaptation. By predicting a constant velocity field across all timesteps, the adaptation task reduces to learning a fixed injection pattern rather than time-varying transformations. Building on this insight, we propose LiteVSR, a minimalist framework that performs VSR using a completely frozen Diffusion Transformer with a lightweight State-Aware Adapter. The adapter employs a dual-stream architecture that extracts static structural cues from the LQ input and dynamic cues from intermediate denoising states, aligning them through time-dependent cross-attention to enable adaptive transition from structural alignment to texture refinement as denoising proceeds. LiteVSR achieves competitive restoration quality with only 11.25% trainable parameters and 12 GPU-hours of training on a single A100, while maintaining fast sampling (down to a single step) compatibility.

Yu Cao, Ziquan Liu, Zhensong Zhang +3

Queen Mary University of London · Huawei Darwin Research Center · Imperial College London

Video Diffusion ModelsDiffusion TransformerVideo Super-ResolutionVideo RestorationParameter-Efficient Fine-Tuning
AskPDF
Jun 6, 2026

How Small Can You Go? LoRA Fine-Tuning 270M-8B Models for Merchant Information Extraction in Financial Transactions

Merchant information extraction turns noisy financial transaction descriptors into structured fields at production scale. Our deployed LoRA-fine-tuned LLaMA3.1-8B reaches 96.95% F1, but its memory and throughput motivate smaller replacements. We evaluate 23 retained fine-tuning runs plus a separately trained production reference, spanning Gemma3 (270M--4B), Qwen3.5 (0.8B--4B), Aya3.35B, and LLaMA3.1-8B across LoRA ranks, prompts, training templates, and serving environments. A rank-8 LLaMA fine-tune reaches 96.75% F1, only 0.20 points below the rank-32 production reference. Qwen3.54B with JSON-Only prompting reaches 96.60% F1 and strict record-level exact match of 91.67%, with a 3.8×3.8\times lower inverse-throughput time estimate than the rank-8 8B model. Qwen3.5~0.8B reaches 94.75% F1, and Qwen Think and Nothink templates differ by less than 0.004 F1. Across 14 Databricks endpoints, mean F1 change from local evaluation is −0.0081-0.0081; Aya is the only family with a 2.7--5.1 point decline. These results show that compact fine-tuned models can preserve most extraction accuracy, but model selection must account for prompt choice, throughput, and serving-stack behavior.

Donghao Huang, Tomas Drietomsky, Benjamin Barrett +1

Research and Development, Mastercard, Arlington, VA, USA · School of Computing and Information Systems, Singapore Management University, Singapore · Economic Intelligence, Mastercard, Prague, Czech Republic

Financial ServicesMemory-Efficient Fine-TuningLLM Fine-TuningDocument Information ExtractionLLM Information ExtractionEfficient Language Model Inference+2
AskPDF
Jun 5, 2026

The Fine-Tuning Trap: Evaluating Negative Transfer and the Role of PEFT in Sub-1B Mathematical Reasoning

Deploying Small Language Models (SLMs) on edge devices requires efficient fine-tuning strategies that adapt models to new tasks without degrading their general capabilities. In this study, we benchmark five sub-1B models (135M-1B) on mathematical reasoning tasks and uncover a critical vulnerability: Full Fine-Tuning (Full FT) actively harms performance in models under 300M parameters, often dropping accuracy below zero-shot baselines. This "negative transfer" makes Parameter-Efficient Fine-Tuning (PEFT) not just an efficiency preference, but a stability requirement. We find that while Low-Rank Adaptation (LoRA) and Weight-Decomposed LoRA (DoRA) perform comparably, their strengths vary by task; DoRA excels in complex reasoning (GSM8K), while LoRA dominates pattern matching (OrcaMath). In particular, Full FT is outperformed by LoRA on aligned models (Qwen2.5-0.5B) and even by simple 5-shot In-Context Learning on the smallest architectures (SmolLM2-135M). Based on these findings, we recommend defaulting to PEFT for all aligned sub-1B models and caution against Full FT for any architecture smaller than 500M parameters to prevent catastrophic forgetting. Reproduction of this work can be found at https://github.com/gulguluu/tiny-slm-finetune-compare.

Rahul Nair, Chun Tao

Fine-TuningSmall Language ModelsNegative TransferParameter-Efficient Fine-Tuning
AskPDF
Jun 4, 2026

TailLoR: Protecting Principal Components in Parameter-Efficient Continual Learning

Parameter-efficient finetuning methods based on spectral decomposition have enabled progress in Continual Learning. In this paper we introduce TailLoR, which utilizes the singular bases U and V of the pre-trained weights as a fixed reference frame to learn a low-rank update applied to the singular value matrix. A soft spectral penalty discourages updates aligned with dominant singular directions, reducing interference while routing fine-grained adaptation into the highly flexible, long-tail spectral coordinates.

Marius Dragoi, Ioana Pintilie, Alexandra Dragomir +2

Bitdefender, Romania

Continual LearningLow-Rank AdaptationParameter-Efficient Fine-TuningContinual LoRA
AskPDF
Jun 4, 2026

Code2LoRA: Hypernetwork-Generated Adapters for Code Language Models under Software Evolution

Code language models need repository-level context to resolve imports, APIs, and project conventions. Existing methods inject this knowledge as long inputs (retrieved through RAG or dependency analysis) or through per-repository fine-tuning and LoRA---costly at repository scale and brittle to evolving codebases. We introduce Code2LoRA, a hypernetwork framework that generates repository-specific LoRA adapters, effectively injecting repository knowledge with zero inference-time token overhead. Code2LoRA supports two usage scenarios: Code2LoRA-Static converts a single repository snapshot into an adapter, suitable for comprehension of stable codebases; while Code2LoRA-Evo maintains an adapter backed by a GRU hidden state updated per code diff, suitable for active development of evolving codebases. To evaluate Code2LoRA against parameter-efficient fine-tuning baselines, we build RepoPeftBench, a benchmark of 604 Python repositories with two tracks: a static track with 40K training and 12K test assertion-completion tasks, and an evolution track with 215K commit-derived training and 87K commit-derived test tasks. On the static track, Code2LoRA-Static achieves 63.8% cross-repo and 66.2% in-repo exact match, matching the per-repository LoRA upper bound; on the evolution track, Code2LoRA-Evo achieves 60.3% cross-repo exact match (+5.2 pp over a single shared LoRA). Code2LoRA's code can be found at https://anonymous.4open.science/r/code2lora-6857; the model checkpoints and RepoPeftBench datasets can be found at https://huggingface.co/code2lora.

Liliana Hotsko, Yinxi Li, Yuntian Deng +1

University of Waterloo

Software EngineeringLLM Fine-TuningCode Language ModelsParameter-Efficient Fine-Tuning
AskPDF
Jun 4, 2026

Rethinking LoRA Memory Through the Lens of KV Cache Compression

Parametric retrieval augmentation encodes document information into lightweight, document-specific modules such as LoRA adapters, reducing the need to include all evidence as input context. However, it remains unclear how this parameter-side memory interacts with context-side memory stored in the KV cache. We study this interaction in document-level question answering by progressively evicting document key-value states and measuring when a document LoRA contributes beyond the retained context. We find that document LoRA adds little when the KV cache is largely intact, but becomes increasingly useful under aggressive compression, recovering 13-21 ROUGE-L points when no document context remains. The gain is largest when the base model encodes the document, and the adapter is applied only during answer generation, suggesting that document LoRA is better understood as decoding-time parametric memory than as a document encoder. Finally, QA-style supervision produces substantially stronger adapters than raw-context next-token-prediction. These results position document LoRA as a complementary memory channel whose value emerges precisely when context-side evidence is scarce.

Chunsheng Zuo, Liaoyaqi Wang, William Jurayj +2

Johns Hopkins University

Retrieval-Augmented GenerationMemory-Augmented Language ModelsKV CachingDocument QAParameter-Efficient Fine-Tuning
AskPDF
Jun 4, 2026

KV-Control: Parameter-Efficient K/V Injection for Trajectory-Controlled Text-to-Motion

Text-conditioned 3D human motion models now synthesize plausible motions from prompts, but practical animation and embodied-agent workflows rarely stop at text: a character may need to follow a sketched root path, hit an end-effector target, or satisfy a multi-joint trajectory while still preserving the gait, style, and intent described by language. This exposes a control trade-off. A trajectory controller should be precise without overwriting the pretrained text-conditioned motion prior, yet existing solutions either duplicate large portions of the generator to regain per-layer control access or move much of the cost to test-time optimization. We introduce KV-Control, a compact attention-side control interface for frozen masked text-to-motion transformers. The key idea is to make geometric constraints available as memory inside self-attention rather than injecting them through a global pose token or enforcing them only at the output side. To support this interface, we co-design a part-tokenized motion substrate and controller: PartVQ learns anatomy-aligned part codebooks, T-Concat exposes each frame--part token as an attention-addressable site, and KV-Control injects control-conditioned key/value memories at every self-attention layer while preserving the pretrained query stream, text cross-attention, FFN, and all backbone weights. The resulting adapter adds only trainable injection parameters atop a shared trajectory encoder, yet tracks root and multi-joint constraints with sub-centimeter accuracy under the inherited refinement protocol while retaining text-conditioned motion quality. KV-Control reframes trajectory conditioning as lightweight memory retrieval, providing a small, precise, and transparent control interface for text-to-motion generation.

Tengjiao Sun, Pengcheng Fang, Xiaoyu Zhan +4

University of Southampton · 2Mogo AI Ltd. · 3Nanjing University

Text-to-Motion GenerationHuman Motion GenerationParameter-Efficient Fine-Tuning
AskPDF
Jun 3, 2026

Dominant-Layer ZO: A Single Layer Dominates Zeroth-Order Fine-Tuning of LLMs

Zeroth-order (ZO) optimization enables memory-efficient fine-tuning of large language models (LLMs) using only forward passes, but it remains unclear how useful adaptation is distributed across layers. In this work, we reveal a surprising phenomenon: ZO fine-tuning is sharply dominated by a single decoding layer. Across multiple LLM families and downstream tasks, fine-tuning this dominant layer alone consistently matches or even exceeds full-model ZO fine-tuning. We further show that the dominant layer is task-agnostic but model-specific, and can be identified before training through a simple inference-only analysis of activation outliers. Specifically, the dominant layer consistently aligns with the first activation-outlier layer in the pre-trained model. To explain this phenomenon, we analyze how perturbation effects propagate under ZO optimization. We find that the dominant layer combines two key properties: high perturbation sensitivity and early placement in the residual stream, allowing perturbation-induced effects to propagate and accumulate through remaining subsequent decoding layers. As a result, this layer produces disproportionately strong and stable optimization signals under forward-only updates. Extensive experiments on LLaMA2-7B and Qwen3-8B across nine benchmarks show that dominant-layer ZO fine-tuning improves average performance over full-model MeZO and LoRA-based ZO fine-tuning while achieving up to 4.52×\times training speedup.

Wanhao Yu, Ziyan Wang, Zheng Wang +7

University of North Carolina at Charlotte · University of Houston · State University of New York at Binghamton +2

Fine-TuningMemory-Efficient Fine-TuningZeroth-Order OptimizationLLM Fine-TuningParameter-Efficient Fine-Tuning
AskPDF
Jun 3, 2026

TaDA: Calibrated Probe Gating for Task-Domain LoRA Merging

Combining a task LoRA adapter with a domain LoRA adapter into a single unified model is a practical yet largely unexplored challenge. Existing methods treat both adapters as symmetric peers, applying uniform weights across all layers. We argue that task and domain adapters exhibit a consistent depth-dependent asymmetry across transformer architectures. Domain dominance increases with layer depth, while shallower layers retain stronger task-relevant signals. Motivated by this observation, we propose TaDA\textbf{TaDA} (Ta\textbf{Ta}sk-D\textbf{D}omain LoRA\textbf{A} Merging), a training-free algorithm that exploits this structure through calibrated probe-guided per-layer gating and per-component subspace-aware merging. The gating assigns individual weights per layer and projection type using a probe signal proved invariant to adapter weight magnitude. The merging discards conflicting singular directions before combining the remaining components. TaDA\textbf{TaDA} produces a standard rank-rr LoRA adapter with zero inference overhead. On six scientific QA benchmarks with Llama-2-7B, TaDA achieves an average accuracy of 0.452, outperforming DARE-TIES by +3.6 percentage points and obtaining the best result on all six benchmarks. On six image classification benchmarks with ViT-L/16, TaDA reaches 85.9% average accuracy, improving over the strongest merging baseline while leading in three of the six individual benchmarks.

Huy Quoc To, Fuyi Li, Guangyan Huang +1

Deakin University, Melbourne, VIC, Australia · Adelaide University, Adelaide, SA, Australia

Adapter TuningLow-Rank AdaptationLanguage Model MergingModel MergingParameter-Efficient Fine-Tuning
AskPDF
Jun 3, 2026

Parameter-Efficient Fine-Tuning with Learnable Rank

Low-Rank Adaptation (LoRA) is a popular parameter-efficient fine-tuning (PEFT) method that restricts weight updates to low-rank adapters, introducing a fixed low-rank inductive bias by optimizing in a low-dimensional subspace. In this work, we question whether a fixed-rank constraint is the most effective inductive bias for parameter-efficient fine-tuning. We introduce Learnable Rank LoRA (LR-LoRA), a PEFT method in which the adapter rank is learned during the training process. Instead of prescribing a uniform rank for all adapter layers, LR-LoRA allows the optimizer to determine the appropriate rank for each layer. Using this approach, we find substantial layer-wise variation in the learned ranks, with the attention and MLP layers in the transformer models exhibiting systematically different rank preferences. Across a range of language understanding and commonsense reasoning benchmarks, LR-LoRA achieves state-of-the-art performance in most settings and consistently outperforms strong PEFT baselines, demonstrating that a learnable rank provides a more flexible and effective inductive bias than fixed-rank adaptations.

Arpit Garg, Simon Lucey, Hemanth Saratchandran

Australian Institute for Machine Learning Adelaide University

Nonlinear LoRALow-Rank AdaptationParameter-Efficient Fine-Tuning
AskPDF
Jun 2, 2026

Where Do We (Not) Need Temporal Context in Low-Resource Video Task Adaptation?

Parameter-efficient finetuning (PEFT) and probing enable adaptation of foundation models using only a small number of trainable parameters. This makes these techniques attractive for video understanding where annotation and computation are expensive. However, video PEFT has focused on adapting image-pretrained models, while standard PEFT methods can also be applied to video representations. These settings are rarely compared and both confine temporal reasoning to a single component of the model, leaving open how temporal context should be distributed across backbone, PEFT, and probe. In this work, we provide a systematic study of model adaptation strategies for video understanding. We evaluate methods across appearance-focused, motion-focused and spatially dense settings, with a particular emphasis on scenarios with limited data where parameter efficiency is most beneficial. Our results provide new insights into PEFT and probing across task types and demonstrate the importance of temporal context allocation for effective video adaptation.

Luc P. J. Sträter, Hazel Doughty

Leiden University

Linear ProbingVision Foundation Model AdaptationParameter-Efficient Fine-TuningTemporal Video Understanding
AskPDF
Jun 2, 2026

Compress then Merge: From Multiple LoRAs into One Low-Rank Adapter

Low-rank adaptation (LoRA) enables parameter-efficient specialization of foundation models, but the proliferation of task-specific adapters fragments capabilities across many adapters, complicating reuse and deployment. We study the problem of merging TT LoRAs into a single rank-rr LoRA, thereby preserving the benefits of low-rank structure. Existing Merge-then-Compress pipelines treat the rank constraint as an afterthought: they merge adapters in the full parameter space, then compress the merged result to rank rr via truncated SVD. However, full-parameter merging may destroy the low-rank structure, making it difficult for subsequent compression to recover an effective rank-rr LoRA. We propose Compress-then-Merge (CtM), a reversed pipeline that enforces the rank-rr bottleneck before merging: CtM computes shared rr-dimensional subspaces using only the LoRA weights to capture cross-adapter common structure, projects each adapter into the shared subspaces to obtain r×rr\times r coordinates, and then applies standard merging rules in this reduced space. CtM guarantees a rank-rr LoRA by construction, avoiding post-hoc truncation, and enables efficient computation in the core space spanned by concatenated LoRA factors. Experiments across multiple models and tasks show that CtM consistently outperforms existing single-LoRA-output baselines while narrowing the performance gap to full-parameter merging methods.

Zhengbao He, Ruiqi Ding, Zhehao Huang +3

Institute of Image Processing and Pattern Recognition, School of Automation and Intelligent Sensing, Shanghai Jiao Tong University, Shanghai, China · Shanghai Key Laboratory of Flexible Medical Robotics, Tongren Hospital, Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, China

Adapter TuningLow-Rank AdaptationLow-Rank CompressionLow-Rank ApproximationModel MergingParameter-Efficient Fine-Tuning
AskPDF
Jun 2, 2026

Message Tuning Outshines Graph Prompt Tuning: A Prismatic Space Perspective

Graph Foundation Models (GFMs), built upon the Pre-training and Adaptation paradigm, have emerged as a research hotspot in graph learning. For GNN-based GFMs, graph prompt tuning has become the prevailing adaptation method for downstream tasks. Although recent methods explain why graph prompt tuning works, how to rigorously measure its adaptation capacity remains an open problem. Addressing this problem is critical for understanding the capability limits of graph prompt tuning and for developing more powerful adaptation methods. In this paper, we propose Prismatic Space Theory (PS-Theory), a novel mathematical framework to quantify the capacity of adaptation methods, while focusing on establishing the upper bound for the adaptation capacity of graph prompt tuning. Building upon the proposed PS-Theory, we further introduce Message Tuning for GFMs (MTG), a lightweight approach that injects a small set of learnable message prototypes into each layer of the GNN backbone to adaptively guide message fusion without updating pre-trained weights. Through our PS-Theory, we prove that the adaptation capacity of MTG can exceed the theoretical upper bound of graph prompt tuning. Extensive experiments demonstrate that MTG consistently outperforms graph prompt baselines across diverse benchmark datasets, providing strong empirical support for our theoretical findings.

Yancheng Chen, Dun Ma, Shuai Zhang +6

Academy of Mathematics and Systems Science, Chinese Academy of Sciences · School of Advanced Interdisciplinary Sciences, University of Chinese Academy of Sciences · Institute of Information Engineering, Chinese Academy of Sciences +4

Prompt TuningGraph Foundation ModelsParameter-Efficient Fine-Tuning
AskPDF
Jun 2, 2026

EURO-5K: When Does Domain Pretraining Matter? Benchmarking Transformers for EU Reporting Obligation Extraction

Extracting reporting obligations from EU legislation is critical for assessing and reducing regulatory reporting burden. However, distinguishing reporting requirements from structurally similar provisions requires specialised legal understanding. Current legal NLP methods lack specialised datasets with clear guidelines and comparative evaluation of extraction paradigms and domain adaptation strategies. We curate EURO-5K, a corpus of sentence-level reporting obligations and challenging negative examples from 136 EU legislative acts. On this dataset, we train and compare discriminative token-classification models (BERT-style) and generative span-extraction models (LLMs), evaluating both full fine-tuning and parameter-efficient QLoRA against baselines (pattern and dependency-based extraction, few-shot prompting). Results show that fully fine-tuned generic and legal BERT models achieve similar performance (0.89 F1), while fine-tuned LLMs match encoder accuracy for sentence-level extraction. Legal pretraining offers only small gains for generative models. In contrast, it is clearly beneficial when adaptation capacity is constrained, as parameter-efficient tuning of Legal-BERT outperforms its generic counterpart. Learning curve analysis demonstrates that legal pretraining accelerates early learning with minimal data. All approaches converge around 3K samples with diminishing returns thereafter, validating dataset sufficiency. Cross-dataset evaluation on two external regulatory corpora shows that our models behave as specialised reporting obligation extractors rather than generic regulatory classifiers. We release EURO-5K, trained models, and an interactive demo with explainability visualizations and structured RDF export. These demonstrate that both paradigms and parameter-efficient training provide practical tools for regulatory compliance automation.

Marios Koniaris, Vasileios Kotronis, Eugenia Giannini +1

Division of Computer Science, School of Electrical and Computer Engineering, National Technical University of Athens, Iroon Polytechniou 9, Zographou Campus, 15780, Athens, Greece. · Department of Humanities Social Sciences and Law, School of Applied Mathematical and Physical Sciences, National Technical University of Athens, Iroon Polytechniou 9, Zographou Campus, 15780, Athens, Greece.

Language Model PretrainingLegal NLPDocument Information ExtractionLegal Document AnalysisParameter-Efficient Fine-Tuning
AskPDF
Jun 1, 2026

CRAM: Centroid-Routing and Adaptive MoE for Multimodal Continual Instruction Tuning

Multimodal Large Language Models (MLLMs) unify heterogeneous vision-language tasks under a shared generative framework via instruction tuning, yet real-world deployment demands continuous capability expansion, making Multimodal Continual Instruction Tuning (MCIT) essential. Existing methods either update all tasks with a shared parameter set or allocate dedicated modules for each new task. Shared updates force heterogeneous tasks to compete, causing forgetting of learned capabilities. Conversely, isolated expansion prevents interference but severely limits parameter efficiency over long task streams. To address this dilemma, we propose CRAM. Specifically, by isolating task-specific patterns into independent modules, CRAM mitigates catastrophic forgetting across tasks. To further boost parameter efficiency, we utilize adaptive-rank instantiation to identify the capability gap between existing expert capability and new task demands, and dynamically allocate only the necessary parameters. To ensure stable reuse among tasks, centroid-guided routing recognizes and activates existing experts' capabilities, while an orthogonality penalty confines new updates to task-specific directions, preventing re-learning general capability. Extensive experiments across diverse benchmarks consistently demonstrate its superiority over existing methods.

Jun-Tao Tang, Zhen-Hao Xie, Yu-Cheng Shi +1

School of Artificial Intelligence, Nanjing University, China · State Key Laboratory of Novel Software Technology, Nanjing University, China

Mixture-of-Experts ModelsMultimodal Instruction TuningParameter-Efficient Fine-Tuning
AskPDF
Jun 1, 2026

On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters

Parameter-efficient fine-tuning (PEFT) is usually treated as a cheaper alternative to full fine-tuning. We study a broader role: small trainable adapters as persistent local state on top of strong shared foundation models. In this framing, the base model provides shared competence while adapters carry instance-specific behavior such as preferences, skills, tool habits, and memory-like updates. We organize the problem around three scaling axes: Scale Up, where stronger shared priors make small local updates more useful; Scale Down, where we study how small adapters can be while remaining reliable; and Scale Out, where many persistent adapted instances coexist. MinT provides one infrastructure example for managing adapter identity, revision, provenance, evaluation, and serving residency. Together, the results suggest that PEFT can be a compact substrate for persistent personal models rather than only a budget substitute for full fine-tuning.

Mind Lab, :, Vin Bo +64

LLM Fine-TuningAdapter TuningPersonalized Language ModelsLanguage ModelingParameter-Efficient Fine-Tuning
AskPDF
Jun 1, 2026

FACT: A Simple and Efficient Framework for Active Finetuning

The main goal of active finetuning is to improve a pretrained model's performance on a specific task or domain by finetuning it with carefully selected informative or challenging data. Previous research has predominantly focused on the active aspect (i.e., data selection) while uniformly employing full finetuning for model adaptation, which inevitably distorts pretrained features due to distribution shift. This issue becomes particularly pronounced when the model size is large relative to the finetuning data quantity, leading to heightened overfitting risks. To address this critical gap, we formally outline the FiAF task that emphasizes systematic exploration of finetuning methodologies in active learning. We propose FACT, a three-phase hierarchical finetuning framework featuring both efficiency and simplicity, specifically designed for active finetuning scenarios. Our comprehensive experiments span: (1) Three major dataset categories encompassing classic (CIFAR10, CIFAR100, ImageNet-1k), imbalanced (CIFAR10-LT, CIFAR100-LT), and fine-grained (StanfordCars, FGVCAircraft) image classification datasets, each evaluated under 3-5 distinct sampling ratios; (2) Diverse pretrained architectures including Convolutional Neural Network (ConvNeXt), Vision Transformer (ViT), and Vision LSTM (ViL) networks; (3) A systematic investigation of frozen feature augmentation (FroFA) strategies. (4) A comprehensive and rigorous analysis of efficiency and generalizability. The results demonstrate significant improvements with strong generalization and robustness. Notably, under low sampling ratios, our framework achieves remarkable performance gains of over 20% on the ViT model for CIFAR10, CIFAR100, and ImageNet-1k benchmarks. This systematic approach establishes new state-of-the-art performance while maintaining parameter efficiency, proving particularly effective when labeled data is scarce.

Wenshuai Xu, You Song, Yuzhuo Cui +3

Active LearningImage ClassificationParameter-Efficient Fine-Tuning
AskPDF
Jun 1, 2026

Parameter-Efficient Fine-Tuning of Large Pretrained Models for Instance Segmentation Tasks

Research and applications in artificial intelligence have recently shifted with the rise of large pretrained models, which deliver state-of-the-art results across numerous tasks. However, the substantial increase in parameters introduces a need for parameter-efficient training strategies. Despite significant advancements, limited research has explored parameter-efficient fine-tuning (PEFT) methods in the context of transformer-based models for instance segmentation. Addressing this gap, this study investigates the effectiveness of PEFT methods, specifically adapters and Low-Rank Adaptation (LoRA), applied to two models across four benchmark datasets. Integrating sequentially arranged adapter modules and applying LoRA to deformable attention--explored here for the first time--achieves competitive performance while fine-tuning only about 1-6% of model parameters, a marked improvement over the 40-55% required in traditional fine-tuning. Key findings indicate that using 2-3 adapters per transformer block offers an optimal balance of performance and efficiency. Furthermore, LoRA, exhibits strong parameter efficiency when applied to deformable attention, and in certain cases surpasses adapter configurations. These results show that the impact of PEFT techniques varies based on dataset complexity and model architecture, underscoring the importance of context-specific tuning. Overall, this work demonstrates the potential of PEFT to enable scalable, customizable, and computationally efficient transfer learning for instance segmentation tasks.

Nermeen Abou Baker, David Rohrschneider, Uwe Handmann

Ruhr West University of applied sciences · Lützowstraße 5, 46236 Bottrop, Germany: Ruhr West University of Applied Sciences

Fine-TuningInstance SegmentationAdapter TuningLow-Rank AdaptationDeformable AttentionParameter-Efficient Fine-Tuning
AskPDF