Machine Learning

Papers filed under cs.LG on arXiv, each one already summarized by Paperlayer. Open any of them to read the summary beside the original PDF, with every point linked to the line, figure, or table it came from.

Search paper metadata (including unsummarized papers)

7,321 to 7,380 of 20,044

Continual Evolution Strategies in Control Tasks
Nicola Pitzalis, Eleni Nisioti, Antonio Carta +2
cs.NEcs.LGarXiv:2608.13600v12026
The Loss Does Not See the Basis, but Adam Does
Devender Singh
cs.LGmath.OCstat.MLarXiv:2608.05136v12026
Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop
Igor Itkin
cs.AIcond-mat.stat-mechcs.CLarXiv:2608.11215v12026
MemSFT: Mitigating Alignment Tax with an External Parametric Memory
Jiarui Wang, Xiang Shi, Jiaqi Cao +8
cs.LGcs.CLarXiv:2607.25614v12026
Kimi Linear: An Expressive, Efficient Attention Architecture
Kimi Team, Yu Zhang, Zongyu Lin +57
cs.CLcs.LGarXiv:2510.26692v22025
ISO: An RLVR-Native Optimization Stack
Hanqing Zhu, Wenyan Cong, Zhizhou Sha +8
cs.LGcs.AIarXiv:2607.19331v12026
Generative Compilation: On-the-Fly Compiler Feedback as AI Generates Code
Niels Mündler-Sasahara, Hristo Venev, Dawn Song +2
cs.PLcs.AIcs.LGarXiv:2607.13921v22026
Summaries:한국어
KronQ: LLM Quantization via Kronecker-Factored Hessian
Donghyun Lee, Yuhang Li, Ruokai Yin +1
cs.LGarXiv:2607.07964v22026
Partially Correlated Verifier Cascades in LLM Harnesses: Concave Log-Odds, Polynomial Reliability, and Blind-Spot Ceilings
Jiangang Han
math.STcs.AIcs.LGarXiv:2607.13918v12026
UI-MOPD: Multi-Platform On-Policy Distillation for Unified GUI Agents
Niu Lian, Tongbo Chen, Zhehao Yu +8
cs.CLcs.AIcs.CVarXiv:2607.04425v22026
Trust Region Policy Distillation
Zhengpeng Xie, Li Lyna Zhang, Zeke Xie +1
cs.LGcs.AIarXiv:2607.04751v12026
MANCE: Manifold Aware Concept Erasure
Matan Avitan, Yoav Goldberg, Yanai Elazar
cs.LGarXiv:2607.03973v12026
Go-with-the-Track: Video Compositing and Motion Control with Point Tracking
Koichi Namekata, Yash Kant, Zhizheng Liu +9
cs.CVcs.LGarXiv:2606.20891v12026
APPO: Agentic Procedural Policy Optimization
Xucong Wang, Ziyu Ma, Yong Wang +5
cs.LGcs.AIarXiv:2606.12384v22026
Test-Time Gradient Guidance of Flow Policies in Reinforcement Learning
Zhiyuan Zhou, Andy Peng, Charles Xu +4
cs.LGcs.AIarXiv:2606.11087v12026
Redesign Mixture-of-Experts Routers with Manifold Power Iteration
Songhao Wu, Ang Lv, Ruobing Xie +1
cs.LGcs.AIcs.CLarXiv:2606.12397v12026
Beyond Uniform Token-Level Trust Region in LLM Reinforcement Learning
Renjie Mao, Xiangxin Zhou, Lvfang Tao +7
cs.LGcs.AIarXiv:2606.10968v22026
Do Coding Agents Deceive Us? Detecting and Preventing Cheating via Capped Evaluation with Randomized Tests
Thanawat Lodkaew, Johannes Ackermann, Soichiro Nishimori +3
cs.LGcs.AIcs.CLarXiv:2606.07379v22026
Beyond Holistic Models: Systematic Component-level Benchmarking of Deep Multivariate Time-Series Forecasting
Shuang Liang, Chaochuan Hou, Xu Yao +4
cs.LGarXiv:2605.26562v12026
OPRD: On-Policy Representation Distillation
Shenzhi Yang, Guangcheng Zhu, Bowen Song +8
cs.LGcs.AIarXiv:2606.06021v42026
Why Muon Outperforms Adam: A Curvature Perspective
Shuche Wang, Fengzhuo Zhang, Jiaxiang Li +2
cs.LGcs.AIarXiv:2606.04662v12026
Two-Fidelity Best-Action Identification for Stochastic Minimax Tree
Peter Chen, Xi Chen
cs.LGcs.AIarXiv:2606.01708v12026
OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification
Yuhang Zhou, Lizhu Zhang, Yifan Wu +5
cs.LGcs.CLarXiv:2606.01476v22026
MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems
Xinle Deng, Ruobin Zhong, Hujin Peng +15
cs.CLcs.AIcs.LGarXiv:2605.28732v32026
Parallax: Parameterized Local Linear Attention for Language Modeling
Yifei Zuo, Dhruv Pai, Zhichen Zeng +3
cs.LGcs.AIcs.CLarXiv:2605.29157v12026
GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models
Xiaohang Tang, Keyue Jiang, Che Liu +4
cs.LGcs.AIarXiv:2605.29398v12026
The Good, the Bad, and the Ugly of Markov Boundary for Tabular Prediction
Shu Wan, Abhinav Gorantla, Huan Liu +1
cs.LGcs.AIstat.MEarXiv:2605.29411v22026
Automatic Construction and Natural-Language Description of Nonparametric Regression Models
James Robert Lloyd, David Duvenaud, Roger Grosse +2
stat.MLcs.LGarXiv:1402.4304v32014
RUBRIC-ARROW: Alternating Pointwise Rubric Reward Modeling for LLM Post-training in Non-verifiable Domains
Haoxiang Jiang, Zihan Dong, Tianci Liu +5
cs.LGcs.CLarXiv:2605.29156v12026
Trust Region Q Adjoint Matching
Yonghoon Dong, Kyungmin Lee, Changyeon Kim +2
cs.LGcs.AIcs.ROarXiv:2605.27079v12026
Pruning and Distilling Mixture-of-Experts into Dense Language Models
Junhyuck Kim, Jihun Yun, Haechan Kim +3
cs.CLcs.AIcs.LGarXiv:2605.28207v22026
Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation
Yuanyi Wang, Su Lu, Yanggan Gu +6
cs.LGarXiv:2605.26844v12026
Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion
Jaihoon Kim, Taehoon Yoon, Prin Phunyaphibarn +3
cs.LGarXiv:2605.23346v12026
When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs
Yifan Zeng, Yiran Wu, Yaolun Zhang +4
cs.AIcs.LGarXiv:2605.24202v22026
Symmetry-Compatible Principle for Optimizer Design: Embeddings, LM Heads, SwiGLU MLPs, and MoE Routers
Tim Tsz-Kit Lau, Weijie Su
math.OCcs.AIcs.LGarXiv:2605.18106v42026
Same Architecture, Different Capacity: Optimizer-Induced Spectral Scaling Laws
Nandan Kumar Jha, Brandon Reagen
cs.LGarXiv:2605.21803v12026
The Distillation Game: Adaptive Attacks & Efficient Defenses
Youssef Allouah, Mahdi Haghifam, Sanmi Koyejo +1
cs.LGcs.AIarXiv:2605.22737v32026
OCTOPUS: Optimized KV Cache for Transformers via Octahedral Parametrization Under optimal Squared error quantization
Mark Boss, Vikram Voleti, Simon Donné +1
cs.LGcs.AIarXiv:2605.21226v12026
optimize_anything: A Universal API for Optimizing any Text Parameter
Lakshya A Agrawal, Donghyun Lee, Shangyin Tan +11
cs.CLcs.AIcs.LGarXiv:2605.19633v12026
CM-EVS: Sparse Panoramic RGB-D-Pose Data for Complete Scene Coverage
Jiale Liu, Jungang Li, Jieming Yu +13
cs.CVcs.GRcs.LGarXiv:2605.15597v12026
Forgetting That Sticks: Quantization-Permanent Unlearning via Circuit Attribution
Saisab Sadhu, Pratinav Seth, Vinay Kumar Sankarapu
cs.LGcs.CLcs.ETarXiv:2605.15138v12026
HodgeCover: Higher-Order Topological Coverage Drives Compression of Sparse Mixture-of-Experts
Tao Zhong, Dongzhe Zheng, Christine Allen-Blanchette
cs.LGcs.AIcs.CLarXiv:2605.13997v12026
Holder Policy Optimisation
Yuxiang Chen, Dingli Liang, Yihang Chen +8
cs.LGcs.AIarXiv:2605.12058v22026
Reliable Chain-of-Thought via Prefix Consistency
Naoto Iwase, Yuki Ichihara, Mohammad Atif Quamar +1
stat.MLcs.CLcs.LGarXiv:2605.07654v12026
Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training
Yuanyi Wang, Yifan Yang, Su Lu +9
cs.LGcs.ITarXiv:2605.09608v12026
G-Zero: Self-Play for Open-Ended Generation from Zero Data
Chengsong Huang, Haolin Liu, Tong Zheng +7
cs.LGcs.AIcs.CLarXiv:2605.09959v12026
Multi-Step-Ahead Time Series Prediction using Multiple-Output Support Vector Regression
Yukun Bao, Tao Xiong, Zhongyi Hu
cs.LGstat.MLarXiv:1401.2504v12014
Sub-JEPA: Subspace Gaussian Regularization for Stable End-to-End World Models
Kai Zhao, Dongliang Nie, Yuchen Lin +4
cs.LGcs.AIarXiv:2605.09241v12026
Rethinking State Tracking in Recurrent Models Through Error Control Dynamics
Jiwan Chung, Heechan Choi, Seon Joo Kim
cs.LGcs.CLarXiv:2605.07755v12026
Mean Mode Screaming: Mean--Variance Split Residuals for 1000-Layer Diffusion Transformers
Pengqi Lu
cs.LGcs.CVarXiv:2605.06169v12026
Addressing Performance Saturation for LLM RL via Precise Entropy Curve Control
Bolian Li, Yifan Wang, Yi Ding +3
cs.LGcs.CLstat.MLarXiv:2604.26326v22026
Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis
Zhisong Qiu, Shuofei Qiao, Kewei Xu +4
cs.CLcs.AIcs.CEarXiv:2604.24198v22026
Sample Selection Using Multi-Task Autoencoders in Federated Learning with Non-IID Data
Emre Ardıç, Yakup Genç
cs.CVcs.LGarXiv:2604.26116v12026
Convergent Evolution: How Different Language Models Learn Similar Number Representations
Deqing Fu, Tianyi Zhou, Mikhail Belkin +2
cs.CLcs.AIcs.LGarXiv:2604.20817v22026
DiPO: Disentangled Perplexity Policy Optimization for Fine-grained Exploration-Exploitation Trade-Off
Xiaofan Li, Ming Yang, Zhiyuan Ma +9
cs.LGarXiv:2604.13902v12026
Reinforcement Learning via Value Gradient Flow
Haoran Xu, Kaiwen Hu, Somayeh Sojoudi +1
cs.LGcs.AIarXiv:2604.14265v12026
Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation
Yecheng Wu, Song Han, Hai Cai
cs.LGcs.AIarXiv:2604.13010v22026
Multi-Domain Riemannian Graph Gluing for Building Graph Foundation Models
Li Sun, Zhenhao Huang, Silei Chen +4
cs.LGarXiv:2603.00618v12026
Steered LLM Activations are Non-Surjective
Aayush Mishra, Daniel Khashabi, Anqi Liu
cs.AIcs.LGarXiv:2604.09839v22026
Tunable Soft Equivariance with Guarantees
Md Ashiqur Rahman, Lim Jun Hao, Jeremiah Jiang +2
cs.CVcs.LGarXiv:2603.26657v12026

Machine Learning

Continual Evolution Strategies in Control Tasks

The Loss Does Not See the Basis, but Adam Does

Poor Man's Agentic Modeling: Simulating Large LLM-Agent Societies on a Laptop

MemSFT: Mitigating Alignment Tax with an External Parametric Memory

Kimi Linear: An Expressive, Efficient Attention Architecture

ISO: An RLVR-Native Optimization Stack

Generative Compilation: On-the-Fly Compiler Feedback as AI Generates Code

KronQ: LLM Quantization via Kronecker-Factored Hessian

Partially Correlated Verifier Cascades in LLM Harnesses: Concave Log-Odds, Polynomial Reliability, and Blind-Spot Ceilings

UI-MOPD: Multi-Platform On-Policy Distillation for Unified GUI Agents

Trust Region Policy Distillation

MANCE: Manifold Aware Concept Erasure

Go-with-the-Track: Video Compositing and Motion Control with Point Tracking

APPO: Agentic Procedural Policy Optimization

Test-Time Gradient Guidance of Flow Policies in Reinforcement Learning

Redesign Mixture-of-Experts Routers with Manifold Power Iteration

Beyond Uniform Token-Level Trust Region in LLM Reinforcement Learning

Do Coding Agents Deceive Us? Detecting and Preventing Cheating via Capped Evaluation with Randomized Tests

Beyond Holistic Models: Systematic Component-level Benchmarking of Deep Multivariate Time-Series Forecasting

OPRD: On-Policy Representation Distillation

Why Muon Outperforms Adam: A Curvature Perspective

Two-Fidelity Best-Action Identification for Stochastic Minimax Tree

OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification

MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems

Parallax: Parameterized Local Linear Attention for Language Modeling

GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models

The Good, the Bad, and the Ugly of Markov Boundary for Tabular Prediction

Automatic Construction and Natural-Language Description of Nonparametric Regression Models

RUBRIC-ARROW: Alternating Pointwise Rubric Reward Modeling for LLM Post-training in Non-verifiable Domains

Trust Region Q Adjoint Matching

Pruning and Distilling Mixture-of-Experts into Dense Language Models

Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation

Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion

When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs

Symmetry-Compatible Principle for Optimizer Design: Embeddings, LM Heads, SwiGLU MLPs, and MoE Routers

Same Architecture, Different Capacity: Optimizer-Induced Spectral Scaling Laws

The Distillation Game: Adaptive Attacks & Efficient Defenses

OCTOPUS: Optimized KV Cache for Transformers via Octahedral Parametrization Under optimal Squared error quantization

optimize_anything: A Universal API for Optimizing any Text Parameter

CM-EVS: Sparse Panoramic RGB-D-Pose Data for Complete Scene Coverage

Forgetting That Sticks: Quantization-Permanent Unlearning via Circuit Attribution

HodgeCover: Higher-Order Topological Coverage Drives Compression of Sparse Mixture-of-Experts

Holder Policy Optimisation

Reliable Chain-of-Thought via Prefix Consistency

Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training

G-Zero: Self-Play for Open-Ended Generation from Zero Data

Multi-Step-Ahead Time Series Prediction using Multiple-Output Support Vector Regression

Sub-JEPA: Subspace Gaussian Regularization for Stable End-to-End World Models

Rethinking State Tracking in Recurrent Models Through Error Control Dynamics

Mean Mode Screaming: Mean--Variance Split Residuals for 1000-Layer Diffusion Transformers

Addressing Performance Saturation for LLM RL via Precise Entropy Curve Control

Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis

Sample Selection Using Multi-Task Autoencoders in Federated Learning with Non-IID Data

Convergent Evolution: How Different Language Models Learn Similar Number Representations

DiPO: Disentangled Perplexity Policy Optimization for Fine-grained Exploration-Exploitation Trade-Off

Reinforcement Learning via Value Gradient Flow

Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation

Multi-Domain Riemannian Graph Gluing for Building Graph Foundation Models

Steered LLM Activations are Non-Surjective

Tunable Soft Equivariance with Guarantees