Hugging Face Trending Papers September 28, 2026

Learn Here, Move Less Elsewhere: Input-Conditioned Plasticity from Retained-Domain Activation Atlases

Read the original on Hugging Face Trending Papers →

The paper introduces ATLAS, a method that uses retained-domain activation atlases to guide task-specific fine‑tuning of language models. By providing local reference centers and directional filters, ATLAS learns a low‑rank residual that adapts to new tasks while preserving existing behavior. Experiments on Qwen3‑8B and other backbones show lower retained‑output KL divergence, fewer rewritten answers, and more stable responses compared to seven baselines.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at Hugging Face Trending Papers.

arXiv Machine Learning

Jun 16

KeepLoRA++: Continual Learning with Layer-Scaled Residual Gradient Adaptation

arXiv:2606. 16256v1 Announce Type: cross Abstract: Continual learning for pre-trained vision-language models requires balancing three competing objectives: retaining pre-trained knowledge, preserving knowledge from a sequence of learned tasks, and maintaining the plasticity to acquire new knowledge.

By Mao-Lin Luo, Yi-Lin Zhang, Zi-Hao Zhou, Yankun Hong, Xialiang Tong, Mingxuan Yuan, Tong Wei, Min-Ling Zhang

llms computer-vision nlp fine-tuning multimodal

arXiv Machine Learning

Sep 24

Attention Routing Stabilizes Early: Working-Set Inference for Recurrent Language Models

The paper investigates how attention dynamics evolve across recurrent depth in language models, finding that attention support stabilizes early while hidden states and outputs take longer. It proposes WISE, a training‑free method that uses full attention in early steps and then reuses the discovered sparse working set for later steps, preserving performance on multi‑hop QA tasks. Experiments show that WISE maintains quality up to 2K context, offers measurable speedups, and highlights the importance of recurrent discovery of attention support.

By Ke Wan, Chen Chen

llms benchmarks

arXiv Computation and Language

3d ago

MetaSteer: Context-Conditioned, nonlinear Steering via Attention-Projection Adaptation

MetaSteer is a new method for steering large language models that learns nonlinear, context-dependent interventions applied to attention projection matrices. Unlike traditional linear, context-independent techniques, MetaSteer adapts its effects based on the input, requiring no linear concept-geometry assumption. Trained once on a pooled preference corpus, it transfers zero‑shot to unseen concepts and out‑of‑distribution contexts, matching or surpassing strong task‑specific baselines on multiple benchmarks and model families.

By Mehdi Jafari, Hao Xue, Flora Salim

llms agents benchmarks safety

arXiv Machine Learning

Jun 8

TALAN: Task-Aligned Latent Adaptation Networks for Targeted Post-Training of Large Language Models

arXiv:2606. 06902v1 Announce Type: new Abstract: Targeted post-training aims to improve reasoning, math, and code without degrading strengths.

By Chengkai Zhang, Ziteng Liu, Junpu Wang, Zeyi Tao, Yang Wang, Sagar Chordia, Qin Huang

llms fine-tuning benchmarks

arXiv Computation and Language

Sep 1

To Copy or Not to Copy: Copying Is Easier to Induce Than Recall

arXiv:2601.12075v2 Announce Type: replace Abstract: Language models used in retrieval-augmented settings must arbitrate between parametric knowledge stored in their weights and contextual information...

By Mehrdad Farahani, Franziska Penzkofer, Richard Johansson

llms rag benchmarks

arXiv AI

Jul 16

Representation-Based Exploration for Language Models: From Test-Time to Post-Training

arXiv:2510. 11686v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) promises to expand the capabilities of language models, but it is unclear if current RL techniques promote the discovery of novel behaviors, or simply sharpen those already present in the base model.

By Jens Tuyls, Dylan J. Foster, Akshay Krishnamurthy, Jordan T. Ash

llms reinforcement-learning

One email a morning, machine-written

Related stories

KeepLoRA++: Continual Learning with Layer-Scaled Residual Gradient Adaptation

Attention Routing Stabilizes Early: Working-Set Inference for Recurrent Language Models

MetaSteer: Context-Conditioned, nonlinear Steering via Attention-Projection Adaptation

TALAN: Task-Aligned Latent Adaptation Networks for Targeted Post-Training of Large Language Models

To Copy or Not to Copy: Copying Is Easier to Induce Than Recall

Representation-Based Exploration for Language Models: From Test-Time to Post-Training