← Back to all news

Fine-tuning and adaptation

LoRA, PEFT, instruction tuning and domain adaptation — adapting a pretrained model without paying to train one.

8,201 stories · RSS feed

arXiv AI
Oct 27

Joint Dynamic Lane Reversal and Variable Speed Limit Control: A Graphormer-Based Approach

The paper proposes a graph-based method for jointly controlling dynamic lane reversal and variable speed limits in traffic networks, ensuring the network remains strongly connected during lane changes. It introduces a transformer model with Global Endpoint Relation-Aware Spatio-Temporal Self-Attention (GERASTSA) that is pretrained on traffic forecasting and fine-tuned with proximal policy optimization. Experiments on a 6×6 grid show the learned policy can make two-thirds of roads one-way during peak hours, reducing daily delay by up to 90% compared to a fixed configuration.

By Hamed Araab, Abbas Ahmadi, Mohammad Hossein Basiri
llmsragfine-tuningsafety
More like this →
arXiv AI
Oct 27

What Can One Sentence Fix? A Failure Taxonomy and a Minimal Language Correction for Frozen VLA Policies

The paper investigates whether a single descriptive sentence can correct failures in frozen vision‑language‑action (VLA) policies. By labeling failures in the LIBERO‑90 benchmark, the authors identify four failure modes, finding that only execution misalignment—where the policy reaches the correct target but misjudges distance or force—can be remedied with a concise human‑generated sentence. This sentence, parsed by an LLM into a bias on the policy’s actions, successfully restores performance on both simulated and real‑world tasks, while automated language models fail to achieve similar gains.

By Owen Kwon, Pablo Ortega-Kral, Arthur Bucker, Jean Oh
llmsroboticsfine-tuningmultimodalsafety
More like this →
arXiv AI
Oct 27

Learning When to Collaborate: Selective Multi-Agent Medical Reasoning via Uncertainty-Aware Routing

The paper introduces SMART‑Med, a local multi‑agent system for medical reasoning that learns when to collaborate instead of always engaging all agents. By using an uncertainty‑driven routing mechanism, it selectively recruits expert agents only for uncertain queries, thereby converting dense collaboration into query‑adaptive coordination. The authors also present a new multimodal dataset, MedVR‑3K, and demonstrate that SMART‑Med matches or surpasses existing multi‑agent approaches while cutting inference time by over 3.8× and token usage by 72%, with a 6.7% average accuracy gain across four VQA benchmarks.

By Jiacheng Hou, Xueliang Cui, Dan Lu, Ruxin Wang
agentsnlpreinforcement-learningfine-tuningmultimodalbenchmarkssafety
More like this →
arXiv AI
Oct 27

EnvLunch: The Environment as a Free Lunch for Learning to Interact

EnvLunch proposes a warm‑up training stage that uses the executable environment itself as a source of verifiable tasks, eliminating the need for human‑written or model‑generated task descriptions. By sampling valid action trajectories, executing them, and training agents to recover equivalent trajectories from final observations, EnvLunch achieves over 9‑point performance gains on downstream tasks and reduces the required annotated examples by 8 to 32 times compared to baselines. This approach also improves fine‑tuning on benchmarks beyond data analysis, offering a more efficient starting point when annotated tasks are scarce.

By Wei Liu, Min Gao, Yali Du, Yulan He
agentsreinforcement-learningfine-tuningbenchmarks
More like this →
arXiv AI
Oct 27

Whispering Wisely: Smart Subset Selection for ASR Fine-Tuning

The paper proposes an embedding‑based subset selection method for fine‑tuning Automatic Speech Recognition models, specifically using Whisper’s pretrained encoder to map audio samples into a fixed‑dimensional space and applying a greedy k‑center algorithm to pick diverse, representative subsets. Experiments on Hebrew and Danish show that k‑center selection outperforms random sampling, especially when the training set is small, and that the chosen subsets exhibit greater lexical diversity. The results suggest that this practical strategy can improve ASR performance under limited data budgets.

By Roi Yozevitch, Or Haim Anidjar
ragfine-tuningmultimodal
More like this →
arXiv AI
Oct 27

Did the Model Use That? Counterfactual Training Boosts Self-Attribution in Language Models

The paper investigates whether language models can self‑attribute the influence of external information on their answers and proposes a counterfactual training method called CREDIT to improve this ability. Experiments show that most state‑of‑the‑art models rarely attribute influential information and often attribute non‑influential information, and that CREDIT significantly enhances attribution in several open models, outperforming other fine‑tuning and prompting approaches. CREDIT also distinguishes between answers that are genuinely changed by user suggestions and those that would have been given regardless, with some transfer to unseen information types and conflicting prompts.

By Zikai Zhou, Jiacheng Sang, Yilin Xu, Yufei Jin, Monica S. Lam
llmsfine-tuning
More like this →
arXiv AI
Oct 27

UTA: Universal Training Adaptive Learning Rate Decay

The paper introduces UTA, a Universal Training Adaptive learning rate decay that treats learning rate decay as a closed‑loop shape‑and‑scale problem rather than a fixed offline trajectory. By formulating the decay in continuous time, UTA adapts to the training state, achieving near‑optimal progression shape and learning‑rate scale. Experiments show consistent improvements on language models (LLaMA 60M‑1B) and vision models (ResNet50, ViT‑B) on ImageNet, with minimal overhead and strong robustness across various hyper‑parameters and fine‑tuning settings.

By Anda Tang, Junjie Wang, Xuanlin Yang, Cong Fang, Zhouchen Lin
llmsfine-tuningbenchmarks
More like this →
arXiv AI
Oct 27

FOCUS: Benchmarking Retinal Model Generalization from Foundation Vision Encoders to Multimodal LLMs

The paper introduces FOCUS, a cross‑dataset benchmark for retinal fundus models that evaluates vision‑only encoders, vision‑language dual encoders, and multimodal large language models across ten public datasets. It measures ranking performance, calibration, subgroup disparities, and image‑quality robustness for tasks such as diabetic retinopathy and glaucomatous optic neuropathy. Results show no single model family dominates, with vision‑only encoders ranking highest on average, while fine‑tuning improves in‑domain performance but varies in external transfer.

By David Restrepo, Chenwei Wu, Luis Filipe Nakayama, Miguel L. Martins, Stergios Christodoulidis, Maria Vakalopoulou, Enzo Ferrante
llmsfine-tuningmultimodalbenchmarkssafety
More like this →