ArXiv TLDR
← All categories

Machine Learning

Papers on learning algorithms, neural networks, deep learning, and optimization.

cs.LG · 1353 papers

Elastic Attention Cores for Scalable Vision Transformers

VECA introduces elastic core-periphery attention for Vision Transformers, achieving linear-time complexity and competitive performance with learned core tokens.

2605.12491May 12, 2026Alan Z. Song, Yinjie Chen, Mu Nan +8

Task-Adaptive Embedding Refinement via Test-time LLM Guidance

This paper introduces an LLM-guided query refinement method that adapts embedding models in real-time for challenging zero-shot search and classification tasks.

2605.12487May 12, 2026Ariel Gera, Shir Ashury-Tahan, Gal Bloch +2

Learning, Fast and Slow: Towards LLMs That Adapt Continually

Fast-Slow Training enables LLMs to adapt continually with improved efficiency and less forgetting by combining fast context and slow parameter updates.

2605.12484May 12, 2026Rishabh Tiwari, Kusha Sareen, Lakshya A Agrawal +6

Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training

A new principle for LM post-training uses sparse rewards for strong teachers and dense distillation for students, outperforming direct sparse RL.

2605.12483May 12, 2026Yuanda Xu, Hejian Sang, Zhengze Zhou +3

MEME: Multi-entity & Evolving Memory Evaluation

MEME is a new benchmark evaluating LLM agents' multi-entity and evolving memory, revealing severe limitations in dependency reasoning.

2605.12477May 12, 2026Seokwon Jung, Alexander Rubinstein, Arnas Uselis +2

Routers Learn the Geometry of Their Experts: Geometric Coupling in Sparse Mixture-of-Experts

This paper reveals a geometric coupling between SMoE routers and experts, explaining how routers learn effective assignment geometry and proposing a coupling-based router.

2605.12476May 12, 2026Sagi Ahrac, Noya Hochwald, Mor Geva

KV-Fold: One-Step KV-Cache Recurrence for Long-Context Inference

KV-Fold enables stable, training-free long-context inference by treating the KV-cache as an accumulator, achieving high fidelity and memory efficiency.

2605.12471May 12, 2026Alireza Nadali, Patrick Cooper, Ashutosh Trivedi +1

Solve the Loop: Attractor Models for Language and Reasoning

Attractor Models introduce a stable, efficient fixed-point refinement method for iterative Transformers, significantly boosting performance in language and reasoning tasks.

2605.12466May 12, 2026Jacob Fein-Ashley, Paria Rashidinejad

High-arity Sample Compression

This paper shows that high-arity sample compression schemes imply high-arity PAC learnability, extending learning theory to product spaces.

2605.12465May 12, 2026Leonardo N. Coregliano, William Opich

Search Your Block Floating Point Scales!

ScaleSearch optimizes Block Floating Point quantization scales by searching for minimal error, significantly improving generative model performance.

2605.12464May 12, 2026Tanmaey Gupta, Hayden Prairie, Xiaoxia Wu +10

Towards Affordable Energy: A Gymnasium Environment for Electric Utility Demand-Response Programs

DR-Gym is a new Gymnasium environment for training RL agents to optimize electric utility demand-response programs, improving grid flexibility and affordability.

2605.12462May 12, 2026Jose E. Aguilar Escamilla, Lingdong Zhou, Xiangqi Zhu +1

A proximal gradient algorithm for composite log-concave sampling

A new proximal gradient algorithm efficiently samples from composite log-concave distributions, matching state-of-the-art for specific cases and extending to broader settings.

2605.12461May 12, 2026Linghai Liu, Sinho Chewi

Multi-Stream LLMs: Unblocking Language Models with Parallel Streams of Thoughts, Inputs and Outputs

Multi-Stream LLMs introduce parallel computation streams to unblock language models, enabling simultaneous reading, thinking, and acting for improved efficiency.

2605.12460May 12, 2026Guinan Su, Yanwu Yang, Xueyan Li +1

TextSeal: A Localized LLM Watermark for Provenance & Distillation Protection

TextSeal is a new LLM watermark using dual-key generation and multi-region localization for robust, distortion-free detection and distillation protection.

2605.12456May 12, 2026Tom Sander, Hongyan Chang, Tomáš Souček +10

Enabling AI-Native Mobility in 6G: A Real-World Dataset for Handover, Beam Management, and Timing Advance

This paper introduces a real-world dataset from a commercial 5G network to enable AI-native mobility, focusing on handover and timing advance.

2605.12453May 12, 2026Mannam Veera Narayana, Rohit Singh, Deepa M. R +1

ORCE: Order-Aware Alignment of Verbalized Confidence in Large Language Models

ORCE improves LLM verbalized confidence by decoupling its estimation from answer generation and using rank-based optimization for better calibration.

2605.12446May 12, 2026Chen Li, Xiaoling Hu, Songzhu Zheng +2

Environment-Adaptive Preference Optimization for Wildfire Prediction

EAPO is a new framework that uses environment-adaptive preference optimization to improve wildfire prediction, especially for rare events and under distribution shifts.

2605.12435May 12, 2026Enyi Jiang, Wu Sun

Learning Minimally Rigid Graphs with High Realization Counts

This paper uses reinforcement learning to discover minimally rigid graphs with record-breaking numbers of realizations, improving bounds for spherical graphs.

2605.12427May 12, 2026Oleksandr Slyvka, Jan Rubeš, Rodrigo Alves +1

ORBIT: Preserving Foundational Language Capabilities in GenRetrieval via Origin-Regulated Merging

ORBIT prevents catastrophic forgetting in GenRetrieval LLMs by regulating weight drift, preserving foundational language capabilities.

2605.12419May 12, 2026Neha Verma, Nikhil Mehta, Shao-Chuan Wang +7

Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space

LLMs update beliefs in a low-dimensional conceptual space, showing in-context learning as trajectories through this space, grounded in structured representations.

2605.12412May 12, 2026Eric Bigelow, Raphaël Sarfati, Daniel Wurgaft +5
PreviousPage 3 of 68Next

📬 Weekly AI Paper Digest

Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.