Skip to yearly menu bar Skip to main content


Workshop Fri, Oct 9, 2026 • 8:30 AM – 6:00 PM PDT Imperial Ballroom B

COLM 2026 Workshop on Efficient Reasoning

Zefan Cai ⋅ Cheng Luo ⋅ Yijiang Li ⋅ Zhenhua Dong ⋅ Haozheng Luo
Project Page

Abstract

Chat is not available.

Schedule

Time zone: Conference local time zone
9:15 AM
Opening Remarks
10:00 AM
Context-Aware RL for Agentic and Multimodal LLMs
Peiyang Xu ⋅ Bangzheng Li ⋅ Sijia Liu ⋅ Karthik Narasimhan ⋅ Pramod Viswanath ⋅ Prateek Mittal ⋅ Xingyu Fu
10:15 AM
Ignorance Is Shared, Knowledge Is Private: Efficiently Mapping Knowledge Boundaries
Owais Aijaz ⋅ Nikhil Ranjan ⋅ Maaz Maqsood ⋅ Ahmer Tabassum ⋅ Haonan Li ⋅ Xuguang Ren ⋅ Zhengzhong Liu ⋅ Preslav Nakov ⋅ Eric Xing
2:30 PM
Express Language Modeling
Albert Gong ⋅ Annabelle M Carrell ⋅ Raaz Dwivedi ⋅ Lester Mackey
2:45 PM
Rebellious Student: Reversing Teacher Signals for Reasoning Exploration with Self-Distilled RLVR
Jeonghye Kim ⋅ Jiwon Jeon ⋅ Dongsheng Li ⋅ Yuqing Yang
3:00 PM
PRESTO: Prefix-Aligned Tree Drafting for Diffusion Speculative Decoding
Zheng Wang ⋅ Zhifan Ye ⋅ Qi Cheng ⋅ Yonggan Fu ⋅ Ziyan Wang ⋅ Feng Zhu ⋅ Haozhe Zhao ⋅ Jan Kautz ⋅ Pavlo Molchanov ⋅ Humphrey Shi ⋅ Minjia Zhang
3:15 PM
Afternoon Break
DARE: Diffusion Language Model Activate Reuse for Efficient Inference
Natalia Frumkin ⋅ Bokun Wang ⋅ Hung-Yueh Chiang ⋅ Chi-Chih Chang ⋅ Mohamed Abdelfattah ⋅ Diana Marculescu
Particles of Thought: Aligning Process Supervision with Test-Time Search
Natasha Butt ⋅ Max Welling ⋅ Jan-Willem van de Meent
Chopthin-Consensus Power Sampling: A Diversity-Preserving Approach to LLM Decoding
Minoo Ahmadi ⋅ Seyedarmin Azizi ⋅ Erfan B Potraghloo ⋅ Mehdi Kamal ⋅ Massoud Pedram
Budget-Calibrated Recovery Routing for Coding Agents
Qijia He ⋅ Jiayi Cheng ⋅ Chenqian Le ⋅ Rui Wang ⋅ Xunmei Liu ⋅ Yixian Chen ⋅ Jie Mei ⋅ Zhihao WANG ⋅ Xupeng Chen ⋅ Yuhuan Chen ⋅ Tao Wang
Decomposing the Delta: What Do Models Actually Learn from Preference Pairs?
Chia-Hsuan Lee ⋅ Mingyang Zhou ⋅ Renkun Ni ⋅ Zelei Cheng ⋅ Sihui Dai ⋅ Supriyo Chakraborty ⋅ Shi-Xiong Zhang ⋅ Sambit Sahu ⋅ William M Campbell
Latent Reasoning with Normalizing Flows
Guancheng Tu ⋅ Xiangjun Fu ⋅ Suhao Yu ⋅ Yao Tang ⋅ Haoqiang Kang ⋅ Lianhui Qin ⋅ Yizhe Zhang ⋅ Jiatao Gu
Solve the Loop: Attractor Models for Language and Reasoning
Jacob Fein-Ashley ⋅ Paria Rashidinejad
Efficient Reinforcement Learning for Long-Horizon Tool-Use Agentic Tasks
Zelei Cheng ⋅ Amritansh Mishra ⋅ Sambit Sahu ⋅ William M Campbell
Inference-Time Scaling Arena: A Common Inference Substrate for Fair and Reproducible Reasoning Evaluation
Seung-seob Lee ⋅ In Gim ⋅ Evan Chen ⋅ Zhiyuan Liu ⋅ Abraam Mankaruse ⋅ Lin Zhong
Leyline: KV-Cache Edit Directives for Efficient Agentic Reasoning
Bole Ma ⋅ Jan Eitzinger ⋅ Harald Koestler ⋅ Gerhard Wellein
RLEM: Reinforcement Learning for External Memory
Dacheng Li ⋅ Shiyi Cao ⋅ Fangzhou Zhao ⋅ Shuo Yuan ⋅ Sumanth Hegde ⋅ Joseph E Gonzalez ⋅ Ion Stoica
Consolidating Rewarded Perturbations for LLM Post-Training
Zheyu Zhang ⋅ Shuo Yang ⋅ Gjergji Kasneci
AvoKV-E: Payload-Aware KV Cache Eviction for Long Reasoning
Han Yu ⋅ Wenhui Zhu ⋅ Xiwen Chen ⋅ Zhipeng Wang ⋅ Hejian Sang ⋅ Han Shi ⋅ Menglin Zhou ⋅ XUANZHAO DONG ⋅ Minzhou Huang ⋅ Rui Cai ⋅ Hao Wang ⋅ Alborz Geramifard
A Reality Check of Language Models as Formalizers on Constraint Satisfaction Problems
Rikhil Amonkar ⋅ Ceyhun E Kayan ⋅ QIMEI LAI ⋅ Ronan Le Bras ⋅ Li Zhang
Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling
Zhen Zhang ⋅ Changyi Yang ⋅ Zijie Xia ⋅ Zhen Yang ⋅ Chengzhi Liu ⋅ Zhaotian Weng ⋅ Yepeng Liu ⋅ Haobo Chen ⋅ Jin Pan ⋅ Chenyang Zhao ⋅ Yuheng Bu ⋅ Alkesh Patel ⋅ Zhe Gan ⋅ Eric Wang
LLM Agents Already Know When to Call Tools - Even Without Reasoning
Chung-En Sun ⋅ Linbo Liu ⋅ Ge Yan ⋅ Zimo Wang ⋅ Tsui-Wei (Lily) Weng
Dynamic Short Convolutions Improve Transformers
Oliver Sieberling ⋅ Bharat Runwal ⋅ Rameswar Panda ⋅ Yoon Kim
QUARTZ: Dequantization-Free LLM Attention via Lookup Tables
Angadjeet Singh ⋅ Parth S Rastogi ⋅ Abhishek Jha ⋅ Mridupawan Das ⋅ Surinder P Singh
BoT-GRPO: Efficient Process-Reward RL for Reasoning via Bag-of-Token Aggregation
Yingxiang Yang ⋅ Weihang Xiao ⋅ Zhunxuan Wang ⋅ Niresh Agarwal
Attention-based Reasoning Outlier Removal for Efficient Reasoning
Haozheng Luo ⋅ Zhuolin Jiang ⋅ Zahid Hasan ⋅ Yan Chen ⋅ Soumalya Sarkar
Brick-DICL: Dynamic In-Context Learning for Automated Brick Schema Classification
Yiyue Qian ⋅ Shinan Zhang ⋅ Huan Song ⋅ Negin Sokhandan ⋅ Hannah R Marlowe ⋅ Diego A Socolinsky
Reasoning Shift: How Context Silently Shortens LLM Reasoning
Gleb Rodionov ⋅ Roman Garipov ⋅ George Yakushev
AdaFlash: Adaptive Speculative Decoding via On-Policy Distilled Diffusion Drafters
Yu-Yang Qian ⋅ Hao-Cong Wu ⋅ Chen Chen ⋅ Jiacheng Sun ⋅ Zhenhua Dong ⋅ Peng Zhao ⋅ Zhi-Hua Zhou
Think Less, Solve More: Efficient Reasoning via Clipped Process Rewards
Sarath Swaminathan ⋅ Payel Das ⋅ Subhajit Chaudhury
Recurrent Key-Value Residual
Zefan Cai ⋅ Cheng Luo ⋅ Hao Tan ⋅ Junjie Hu
MEMGYM: a Long-Horizon Memory Environment for LLM Agents
Wujiang Xu ⋅ Yu Wang ⋅ Kai Mei ⋅ Kaiqu Liang ⋅ Zhenting Wang ⋅ Mingyu Jin ⋅ Han Zhang ⋅ Shi-Xiong Zhang ⋅ Wenyue Hua ⋅ Sambit Sahu ⋅ Dimitris N Metaxas
SOFAI-DT: Return-Conditioned Control for Efficient Verified Reasoning
Rahul Maity ⋅ Vedant Khandelwal ⋅ Amit Sheth
TraceNAS: Zero-shot Search for LLM Pruning via Gradient Trace Correlation
Prajna G. Malettira ⋅ Manish Nagaraj ⋅ Arjun Roy ⋅ Shubham Negi ⋅ Kaushik Roy
Prefix sliding for efficient test-time scaling
Niklas Muennighoff ⋅ Zhengyang Wang ⋅ Zeyi Chen ⋅ Weijia Shi ⋅ Binyuan Hui ⋅ John Yang ⋅ Dapeng Jiang ⋅ Mika Senghaas ⋅ Fares Obeid ⋅ Johannes Hagemann ⋅ Sami Jaghouar ⋅ Ludwig Schmidt ⋅ Percy Liang ⋅ Jason Wei ⋅ Andrew Y Ng ⋅ Luke Zettlemoyer ⋅ Yejin Choi ⋅ Mike Lewis
VLM Token Compression through Task-Adaptive Predictor Selection: A System-aware Co-Design Analysis and Practice
Yiran Shi ⋅ Ding Ni ⋅ Junming J CHEN ⋅ Tianchen Zhao ⋅ Yaqi Yan ⋅ Yu Wang
Reward-Gated On-Policy Distillation
Mohammad Sadegh Akhondzadeh ⋅ Vijay Lingam ⋅ Atula Tejaswi ⋅ Chanakya Ekbote ⋅ Sujay Sanghavi ⋅ Aleksandar Bojchevski
Knowledge‑Distilled Memory Editing for Plug‑and‑Play LLM Alignment
Haozheng Luo ⋅ Jiahao Yu ⋅ Wenxin Zhang ⋅ Jialong Li ⋅ Chenghao Qiu ⋅ Yimin Wang ⋅ Eric H Jiang ⋅ Ching-Yuen Huang ⋅ Chenwei Xu ⋅ Jerry Y Hu ⋅ Yan Chen ⋅ Binghui Wang ⋅ Xinyu Xing ⋅ Han Liu
Context Pruning for Coding Agents via Latent Multi-Rubric Reasoning
Jingjing Wang ⋅ Xiwen Chen ⋅ Wenhui Zhu ⋅ Huayu Li ⋅ ZhengXiao He ⋅ Feiyang Cai ⋅ Ana S Carreon-Rascon ⋅ XUANZHAO DONG ⋅ Feng Luo
S2D2: Fast Decoding for Diffusion LLMs via Training-Free Self-Speculation
Ligong Han ⋅ Hao Wang ⋅ Han Gao ⋅ Kai Xu ⋅ Akash Srivastava
Algorithmic Primitives of Theory of Mind Benchmarks
Qiutong Yi ⋅ Wang Bill Zhu ⋅ Robin Jia ⋅ Jesse Thomason
RebaseKV: Hiding Prefill Latency in Multi-Agent LLMs via Optimistic Cache Alignment
Jiayu Qin ⋅ Ankith M S ⋅ Harpreet Kaur ⋅ Kevin Zhai ⋅ Sumit Negi ⋅ Changyou Chen
On Asymmetric Optimization of Reasoning and Perception in Vision-Language Model Post-Training
Xueqing Wu ⋅ Yu-Chi Lin ⋅ Kai-Wei Chang ⋅ Nanyun (Violet) Peng
Before You Scale: The Cost of Supervision Mismatch in PRM Distillation
Saksham Kapoor ⋅ Henry Tran ⋅ Edward L Lip ⋅ Charlotte Le
Inject a Turn, Not a Token: Learning Where to Redirect LLM Reasoning for Self-Correction
Jiwon Jeon ⋅ Myungsik Cho ⋅ Jeonghye Kim ⋅ Youngchul Sung
SEGRA: A Structured Experience Guided Reasoning Agent for Property Graph Question Answering
Saiyue Lyu ⋅ Mariam Dundua ⋅ Vishaal Kapoor ⋅ Sarthak Ahuja ⋅ Neda Kordjazi ⋅ Evren Yortucboylu ⋅ Harsh Amin ⋅ Rebecca Steinert
Test-Time Generation
Zefan Cai ⋅ Hao Tan ⋅ Yuchen Zhu ⋅ Tianyuan Zhang ⋅ Sai Bi ⋅ Haoyi Qiu ⋅ Yicong Hong ⋅ Jason Kuen ⋅ Haoliang Wang ⋅ Haozhe Zhao ⋅ Ruiyi Zhang ⋅ Ryan Rossi ⋅ Wanrong Zhu ⋅ Molei Tao ⋅ Wen Xiao ⋅ Rajiv Jain ⋅ Junjie Hu ⋅ Jiuxiang Gu
ADOPD 2026: From Page Decomposition to Grounded Document Reasoning
Sichen Zhu ⋅ Yuchen Zhu ⋅ Jason Kuen ⋅ Wanrong Zhu ⋅ Jing Shi ⋅ Xuan Shen ⋅ Quanyi Wang ⋅ Yiwei Wang ⋅ Yujun Cai ⋅ Bing Shuai ⋅ Qin ZHANG ⋅ Yongxin Chen ⋅ Shilong Liu ⋅ Molei Tao ⋅ Jiuxiang Gu
SCaTR: Simple Calibrated Test-Time Ranking
Divya Shyamal ⋅ Marta Knežević ⋅ Lan Tran ⋅ Chanakya Ekbote ⋅ Vijay Lingam ⋅ Paul Pu Liang
Latent Reasoning with Supervised Thinking States
Ido Amos ⋅ Avi Caciularu ⋅ Jonathan Herzig ⋅ Mor Geva ⋅ Amir Globerson ⋅ Lior Shani ⋅ Idan Szpektor
Clipping-Free Policy Optimization for Large Language Models
Ömer Çağatan ⋅ Baris Akgun ⋅ Gözde G Şahin ⋅ Xuandong Zhao
Multi-Head Attention Residuals
Cheng Luo ⋅ Zefan Cai ⋅ Junjie Hu
ReWork: Efficient Iterative Reasoning with Working Memory
Haoyu He ⋅ Yong Cao ⋅ Andreas Geiger
Ignorance Is Shared, Knowledge Is Private: Efficiently Mapping Knowledge Boundaries
Owais Aijaz ⋅ Nikhil Ranjan ⋅ Maaz Maqsood ⋅ Ahmer Tabassum ⋅ Haonan Li ⋅ Xuguang Ren ⋅ Zhengzhong Liu ⋅ Eric Xing
Fully-Recurrent Test-Time Training
Zefan Cai ⋅ Hao Tan ⋅ Junjie Hu
Librarian Agents: Anticipatory Filesystem Construction via Reinforcement Learning
Anikait Singh ⋅ Teresa Zhang ⋅ Yoonho Lee ⋅ Roshen S Nair ⋅ Aviral Kumar ⋅ Chelsea Finn
Minimum Sufficient Reasoning: How Much Chain-of-Thought Does Training Data Need?
Maaz Maqsood ⋅ Nikhil Ranjan ⋅ Owais Aijaz ⋅ Ahmer Tabassum ⋅ Haonan Li ⋅ Xuguang Ren ⋅ Zhengzhong Liu ⋅ Eric Xing
Gaze Attention: Selective Visual Attention for Multimodal Large Language Models
Junha Song ⋅ Byeongho Heo ⋅ Geonmo Gu ⋅ Jaegul Choo ⋅ Dongyoon Han ⋅ Sangdoo Yun
AMVICC: A Novel Benchmark for Cross-Modal Failure Mode Profiling for VLMs and IGMs
Pranay Goel ⋅ Aahana Basappa ⋅ Anusri Karra ⋅ Anish Karra ⋅ ⋅ Kevin Zhu
UniPath: Adaptive Coordination of Understanding and Generation for Unified Multimodal Reasoning
Hayes Bai ⋅ Yinyi Luo ⋅ Wenwen Wang ⋅ Qingsong Wen ⋅ Jindong Wang
ConstOptBench: Evaluating LLMs on Verifiable Constrained Optimization Reasoning without Tool Use
Derek Flippo ⋅ Lauren Cho ⋅ Pratima Nallapareddy ⋅ Safwan Rahman ⋅ Yuheng Bu ⋅ Yepeng Liu
When Hidden States Drift: Can KV Caches Rescue Long-Range Speculative Decoding?
Tianyu Liu ⋅ Yuhao Shen ⋅ Xinyi Hu ⋅ Zhang Baolin ⋅ Hengxin Zhang ⋅ Jun Dai ⋅ Jun Zhang ⋅ Shuang Ge ⋅ Chen Lei ⋅ Yue Li ⋅ Mingcheng Wan ⋅ Xiaoyan Sun ⋅ Xiao Sun
OpenResearcher: A Fully Open Pipeline for Long-Horizon Deep Research Trajectory Synthesis
Zhuofeng Li ⋅ Dongfu Jiang ⋅ Xueguang Ma ⋅ Haoxiang Zhang ⋅ Ping Nie ⋅ Yuyu Zhang ⋅ Kai Zou ⋅ Jianwen Xie ⋅ Yu Zhang ⋅ Wenhu Chen
Optimal Context Stopping for Large Language Models
Yufa Zhou ⋅ Weili Cao ⋅ Haoze Liu ⋅ Anru Zhang
Your Agent is More Brittle Than You Think: Efficiently Mitigating Indirect Injections in Agentic LLMs
Wenhui Zhu ⋅ XUANZHAO DONG ⋅ Xiwen Chen ⋅ Rui Cai ⋅ Peijie Qiu ⋅ Zhipeng Wang ⋅ Oana Frunza ⋅ Yueyue Deng ⋅ Shao Tang ⋅ Jindong Gu ⋅ Yalin Wang
Characterizing Model-Native Skills
Feiyang Kang ⋅ Mahavir Dabas ⋅ Myeongseob Ko ⋅ Ruoxi Jia
Express Language Modeling
Albert Gong ⋅ Annabelle M Carrell ⋅ Raaz Dwivedi ⋅ Lester Mackey
Stepwise Penalization for Length-Efficient Chain-of-Thought Reasoning
Xintong Li ⋅ Sha Li ⋅ Rongmei Lin ⋅ Hongye Jin ⋅ Linwei Li ⋅ Hejie Cui ⋅ Sarah Zhang ⋅ Chia-Yuan Chang ⋅ Kewei Cheng ⋅ Yuwei Zhang ⋅ Qingyu Yin ⋅ Jingbo Shang ⋅ Bing Yin
From Static to Systematic: A Survey on Efficient Inference Across LLM, RL, and Agentic Eras
Zhenglin Wang ⋅ Sichu Liang ⋅ Congzhi Zhang ⋅ Bin Zhang ⋅ Jiahao Yuan ⋅ Haowen Sun ⋅ Ziyang Ma ⋅ Mengfei Zhang ⋅ Yilong Lai ⋅ Jialong Wu ⋅ Zhenyang Li ⋅ WANG Zongwei ⋅ Tong Liu
Message Passing Enables Efficient Reasoning
Xuecheng Liu ⋅ Daman Arora ⋅ Gokul Swamy ⋅ Andrea Zanette
QuasiMoTTo: Quasi-Monte Carlo Test-Time Scaling
Michael Y Li ⋅ Anthony Zhan ⋅ Kanishk Gandhi ⋅ Noah Goodman ⋅ Emily Fox
AriadneMem: Threading the Maze of Lifelong Memory for LLM Agents
Wenhui Zhu ⋅ Xiwen Chen ⋅ Zhipeng Wang ⋅ Jingjing Wang ⋅ XUANZHAO DONG ⋅ Rui Cai ⋅ Hejian Sang ⋅ Hao Wang ⋅ Peijie Qiu ⋅ Yueyue Deng ⋅ Minzhou Huang ⋅ Prayag Tiwari ⋅ Brendan H Rappazzo ⋅ Yalin Wang
MED-EVAL: Clinical Fact Alignment for Efficient Medical Reasoning
Rishit Agarwal ⋅ Vedant Khandelwal ⋅ Amit Sheth
Claim-Level Reliability Assessment for Efficient Test-Time Reasoning
Sen Xu ⋅ Wei Wang ⋅ Shixi Liu ⋅ Jixin Min ⋅ Yingwei Dai ⋅ Zhibin Yin ⋅ Chen Yirong ⋅ Junlin Zhang
Rethinking On-Policy Self-Distillation for Thinking Models
Simran Kaur ⋅ Narutatsu Ri ⋅ Yinghui He ⋅ Liam H Fowl ⋅ Sanjeev Arora
From Projection Geometry to Attention Fidelity: Explaining Cross-Layer KV Prediction
Tinuade Adeleke ⋅ William Stella ⋅ Arya Srikar Reddy ⋅ Ruizhe Li
Agentic Test-Time Scaling for WebAgents
Nicholas Lee ⋅ Lutfi E Erdogan ⋅ Chris John ⋅ Surya Krishnapillai ⋅ Michael W Mahoney ⋅ Kurt Keutzer ⋅ Amir Gholami
You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectories
Zhepei Wei ⋅ Xinyu Zhu ⋅ Wei-Lin Chen ⋅ Chengsong Huang ⋅ Jiaxin Huang ⋅ Yu Meng
From Refusal to Fabrication: An Affective Cause of Hallucination That Bypasses the Honesty Gate
Akshat Agarwal ⋅ Brayden Feng ⋅ William Qian ⋅ Ruizhe Li ⋅ Aditya Shah ⋅ Vasu Sharma ⋅ Maheep Chaudhary
Layer-Parallel Inference Reduces Encrypted Nonlinear Depth in Transformers
Ligong Han ⋅ Kai Xu ⋅ Hao Wang ⋅ Ruijiang Gao ⋅ Han Gao ⋅ Akash Srivastava
Distribution Corrected Offline Data Distillation for Large Language Models
Yumeng Zhang ⋅ Zhengbang Yang ⋅ Yevin N Goonatilake ⋅ Zhuangdi Zhu
Attention Sinks and Outliers in Attention Residuals
Haozheng Luo ⋅ Haoran Dai ⋅ Shaoyang Zhang ⋅ Xi Chen ⋅ Yijiang Li ⋅ Ching-Yuen Huang ⋅ Chenghao Qiu ⋅ Chenwei Xu ⋅ Zhenyu Pan ⋅ Jiakun Fan ⋅ Haotian Zhang ⋅ Binghui Wang ⋅ Yan Chen
Multiplex Thinking: Reasoning via Token-wise Branch-and-Merge
Yao Tang ⋅ Li Dong ⋅ Yaru Hao ⋅ Qingxiu Dong ⋅ Furu Wei ⋅ Jiatao Gu
PlanForge: Synthesizing Planners from Natural Language with an LLM-Driven Architect--Builder--Runner Pipeline
Sarvesh Baskar ⋅ Raviteja Bommireddy ⋅ Saksham K Sharma ⋅ Harsha Kokel ⋅ Manas Gaur ⋅ Michael Katz ⋅ Shirin Sohrabi ⋅ Kavitha Srinivas
Early Stopping for Large Reasoning Models via Confidence Dynamics
Parsa Hosseini ⋅ Sumit Nawathe ⋅ Mahdi Salmani ⋅ Meisam Razaviyayn ⋅ Soheil Feizi
Can RLVR Recover Reasoning Quality Missing from Mid-Training?
Tianyi Zhang ⋅ Stella Li ⋅ Yulia Tsvetkov
Prompt-Cooperative KV Cache Reuse for Efficient Multi-Agent LLM Inference
Jiayu Qin ⋅ Ankith M S ⋅ Harpreet Kaur ⋅ Kevin Zhai ⋅ Sumit Negi ⋅ Changyou Chen
Value-Aware Stochastic KV Cache Eviction for Reasoning Models
Ting-Yun Chang ⋅ Harvey Yiyun Fu ⋅ Deqing Fu ⋅ Chenghao Yang ⋅ Jesse Thomason ⋅ Robin Jia
Know When to Stop: Segment-Level Credit Assignment for Reducing Overthinking
Chia-Hsuan Lee ⋅ Sihui Dai ⋅ Mingyang Zhou ⋅ Isha Slavin ⋅ Shi-Xiong Zhang ⋅ Sambit Sahu ⋅ William M Campbell
How Should Code LLMs Learn Execution? A Study of Trace Representations
Roman Machacek ⋅ Jonas Spieler ⋅ Timo Kehrer
Arbitrage: Efficient Reasoning via Advantage-Aware Speculation
Monish Maheswaran ⋅ Rishabh Tiwari ⋅ Yuezhou Hu ⋅ Kerem Dilmen ⋅ Coleman R Hooper ⋅ Haocheng Xi ⋅ Nicholas Lee ⋅ Mehrdad Farajtabar ⋅ Michael W Mahoney ⋅ Kurt Keutzer ⋅ Amir Gholami
Reasoning Ability Is Not Reasoning Economics: Benchmarking Sequential Budget Allocation in LLMs
Yashaswi Sharma ⋅ Prabhu Pugalenthi ⋅ Muyan Weng ⋅ Defu Cao ⋅ Yan Liu
Planned Test-Time Scaling with Coordinated Reasoning Paths
Xueqing Wu ⋅ Langxing Bai ⋅ Hritik Bansal ⋅ Po-Nien Kung ⋅ Shuo Li ⋅ Hao Liu ⋅ Nanyun (Violet) Peng ⋅ Kai-Wei Chang
Context-Aware RL for Agentic and Multimodal LLMs
Peiyang Xu ⋅ Bangzheng Li ⋅ Sijia Liu ⋅ Karthik Narasimhan ⋅ Pramod Viswanath ⋅ Prateek Mittal ⋅ Xingyu Fu
Understanding Stochastic Reasoning Dynamics in LLM Judges through Markov Models
Giuliano Lorenzoni ⋅ Paulo Alencar ⋅ Donald Cowan
Process Reward Informed Tree Rollout for Effective Multi-Turn RL
Xintong Li ⋅ Sha Li ⋅ Yuwei Zhang ⋅ Changlong Yu ⋅ Rongmei Lin ⋅ Hongye Jin ⋅ Shuyi Guan ⋅ Xin Liu ⋅ Linwei Li ⋅ Qingyu Yin ⋅ Jingbo Shang
Safety Lost in Translation: Usable Yield and the Cost of Thinking in Safety-Data Localization
Ahmer Tabassum ⋅ Nikhil Ranjan ⋅ Maaz Maqsood ⋅ Owais Aijaz ⋅ Haonan Li ⋅ Xuguang Ren ⋅ Zhengzhong Liu ⋅ Preslav Nakov ⋅ Eric Xing
CoMem: Reusing Transformer Depth across Queries with Persistent Intermediate Residuals
Hanzuo Liu ⋅ Xuan Qi ⋅ Chunyu Liu ⋅ Yulong Wang ⋅ Rayying ⋅ key ⋅ Alex Lamb ⋅ Mingyu Gao
Beyond Reward Engineering: A Data Recipe for Long-Context Reinforcement Learning
Xiaoyue Xu ⋅ Sikui Zhang ⋅ xiaorong wang ⋅ Xu Han ⋅ Chaojun Xiao
FitText: Evolving Agent Tool Ecologies via Memetic Retrieval
Kyle Zheng ⋅ Han Zhang ⋅ Renliang Sun ⋅ Chenchen Ye ⋅ Wei Wang
Playing with Fire: What Transfers When RL Trains a Language Agent?
Mahesh Ramesh ⋅ Kaousheik Jayakumar ⋅ Hemanth Ram ⋅ Pavan Thodima ⋅ Ramani Duraiswami ⋅ Dinesh Manocha ⋅ Aniket Rege ⋅ Emmanouil-Vasileios Vlatakis-Gkaragkounis
Efficient Rewarding of Agent Plans Without LLM-as-Judge: Goal Drift Score for GRPO
Feyzi A Salihoğlu ⋅ Yağız E Dalar ⋅ Nedim M Sezer ⋅ Ahmet Rıfat Öztürk ⋅ Ömer F Aksoy
Learning to Generalize Recursively: Structural Insights from the Tower of Hanoi
Junyi Liao ⋅ Ethan X Fang ⋅ Zhuoran Yang ⋅ Vahid Tarokh
Step-Level Optimization for Efficient Computer-Use Agents
Jinbiao Wei ⋅ Kangqi Ni ⋅ Yilun Zhao ⋅ Guo Gan ⋅ Arman Cohan
Playing Psychic: Using Thought Trees to Predict Reasoning Models Accuracy on Coding Tasks
Jiaxin Fang ⋅ Runyuan He ⋅ Sahil Bhatia ⋅ Neel Gajare ⋅ Alvin Cheung
Efficient Reasoning with Flow Language Models
Hanru Bai ⋅ Faissal IZERMINE ⋅ Oscar Davis ⋅ T. Konstantin Rusch
Self-Speculative Biased Decoding for Faster Re-translation
Linxiao Zeng ⋅ Haoyun Deng ⋅ Kangyuan Shu ⋅ Shizhen Wang
TextReg: Mitigating Prompt Distributional Overfitting via Regularized Text-Space Optimization
Lucheng Fu ⋅ Ye Yu ⋅ Yiyang Wang ⋅ Yiqiao Jin ⋅ Haibo Jin ⋅ B. Aditya Prakash ⋅ Haohan Wang
PRESTO: Prefix-Aligned Tree Drafting for Diffusion Speculative Decoding
Zheng Wang ⋅ Zhifan Ye ⋅ Qi Cheng ⋅ Yonggan Fu ⋅ Ziyan Wang ⋅ Feng Zhu ⋅ Haozhe Zhao ⋅ Jan Kautz ⋅ Pavlo Molchanov ⋅ Humphrey Shi ⋅ Minjia Zhang
Last Layer Key-Value Cache is All You Need
Zefan Cai ⋅ Sai Bi ⋅ Lin Zhang ⋅ Yicong Hong ⋅ Chongjian GE ⋅ Xun Huang ⋅ Jiuxiang Gu ⋅ Junjie Hu ⋅ Hao Tan
Test-Time Training with Next-Token Prediction
Xuan Ouyang ⋅ Zefan Cai ⋅ Junjie Hu
Reason Wide, Not Deep: Amortizing the Reasoning Premium into Distilled Skills
Agamdeep Singh ⋅ Srishti Gautam ⋅ Priyanshu Gupta ⋅ Nikita Mehrotra ⋅ Tanmay Bakshi ⋅ Sumit Gulwani
CAST: Contrastive Adjudication via Selective Triggering for Efficient LLM Reasoning
Krishna Senthilkumar ⋅ Archana Vaidheeswaran ⋅ Shivank Garg
Stubborn or Sycophantic? GEPA-Evolved Prompts Under Pressure
HanRui Zhang ⋅ Ashton Chew ⋅ Ryan Soe ⋅ Ahmed Taha ⋅ ⋅ Aditya Shah ⋅ Maheep Chaudhary
Test-Time Training Normalization Done Right
Zefan Cai ⋅ Hao Tan ⋅ Junjie Hu
Universal Test-Time Training
Zefan Cai ⋅ Qinzhe Hu ⋅ Ziqiao Ma ⋅ Hao Tan ⋅ Junjie Hu
Test-Time Scaling via Error Localization
Rajiv Shailesh Chitale ⋅ Rahul Madhavan ⋅ Taneesh Gupta ⋅ Deepanway Ghosal ⋅ Aravindan Raghuveer
SUPERNOVA: Eliciting General Reasoning in LLMs with Reinforcement Learning on Natural Instructions
Ashima Suvarna ⋅ Kendrick Phan ⋅ Mehrab Beikzadeh ⋅ Hritik Bansal ⋅ Saadia Gabriel
Beyond Hidden Chain-of-Thought: Mechanistic Evidence for Distributed Latent Reasoning
Aaron Ren ⋅ Ahmed Mohammed ⋅ Viet M Nguyen ⋅ Bach Nguyen ⋅ Cole Blondin
SODA: Semi On-Policy Black-Box Distillation for Large Language Models
Xiwen Chen ⋅ Jingjing Wang ⋅ Wenhui Zhu ⋅ Peijie Qiu ⋅ XUANZHAO DONG ⋅ Yueyue Deng ⋅ Hejian Sang ⋅ Zhipeng Wang ⋅ Alborz Geramifard ⋅ Feng Luo
Chain-of-Experience for Continual LLM Improvement
Haoqin Tu ⋅ Yunhao Fang ⋅ Yizhong Wang ⋅ Cihang Xie ⋅ Shen Yan
A Model with No Head and Many Thoughts
Nikita Koriagin ⋅ Yaroslav Aksenov ⋅ George Bredis ⋅ Gleb Gerasimov ⋅ Nikita Balagansky ⋅ Daniil Gavrilov
Vector Policy Optimization: Training for Diversity Improves Test-Time Search
Ryan Bahlous-Boldi ⋅ Isha Puri ⋅ Idan Shenfeld ⋅ Akarsh Kumar ⋅ Mehul Damani ⋅ Sebastian Risi ⋅ Omar Khattab ⋅ Zhang-Wei Hong ⋅ Pulkit Agrawal
MMGR: Multi-Modal Generative Reasoning Benchmark and Evaluation
Zefan Cai ⋅ Haoyi Qiu ⋅ Tianyi Ma ⋅ Haozhe Zhao ⋅ Gengze Zhou ⋅ Tingting Liao ⋅ Xinyan Velocity Yu ⋅ Kung-Hsiang Huang ⋅ Ke Wan ⋅ Shawn Lin ⋅ Parisa Kordjamshidi ⋅ Minjia Zhang ⋅ Wen Xiao ⋅ Jiuxiang Gu ⋅ Nanyun (Violet) Peng ⋅ Junjie Hu
CRAFT: Reinforcement Learning from What Models Think, Not What They Say
Haozheng Luo ⋅ Yimin Wang ⋅ Haoran Dai ⋅ Jiahao Yu ⋅ Binghui Wang ⋅ Yan Chen
Post-training Trade-offs Under Limited Compute Budgets
Shougan Gu ⋅ Michael Ogezi ⋅ Freda Shi
Neural Garbage Collection: Learning to Forget while Learning to Reason
Michael Y Li ⋅ Jubayer I Hamid ⋅ Emily Fox ⋅ Noah Goodman
DARE: Diffusion Language Model Activate Reuse for Efficient Inference
Natalia Frumkin ⋅ Bokun Wang ⋅ Hung-Yueh Chiang ⋅ Chi-Chih Chang ⋅ Mohamed Abdelfattah ⋅ Diana Marculescu
ARCHER: Adaptive Recovery Routing with Conformal-risk Calibration for Coding Agents
Qijia He ⋅ Jiayi Cheng ⋅ Chenqian Le ⋅ Rui Wang ⋅ Xunmei Liu ⋅ Yixian Chen ⋅ Jie Mei ⋅ Zhihao WANG ⋅ Xupeng Chen ⋅ Yuhuan Chen ⋅ Tao Wang
Leyline: KV-Cache Edit Directives for Efficient Agentic Reasoning
Bole Ma ⋅ Jan Eitzinger ⋅ Harald Koestler ⋅ Gerhard Wellein
A Reality Check of Language Models as Formalizers on Constraint Satisfaction Problems
Rikhil Amonkar ⋅ Ceyhun E Kayan ⋅ QIMEI LAI ⋅ Ronan Le Bras ⋅ Li Zhang
Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling
Zhen Zhang ⋅ Changyi Yang ⋅ Zijie Xia ⋅ Zhen Yang ⋅ Chengzhi Liu ⋅ Zhaotian Weng ⋅ Yepeng Liu ⋅ Haobo Chen ⋅ Jin Pan ⋅ Chenyang Zhao ⋅ Yuheng Bu ⋅ Alkesh Patel ⋅ Zhe Gan ⋅ Eric Wang
Attention-based Reasoning Outlier Removal for Efficient Reasoning
Haozheng Luo ⋅ Zhuolin Jiang ⋅ Zahid Hasan ⋅ Yan Chen ⋅ Soumalya Sarkar
Reasoning Shift: How Context Silently Shortens LLM Reasoning
Gleb Rodionov ⋅ Roman Garipov ⋅ George Yakushev
AdaFlash: Adaptive Speculative Decoding via On-Policy Distilled Diffusion Drafters
Yu-Yang Qian ⋅ Hao-Cong Wu ⋅ Chen Chen ⋅ Jiacheng Sun ⋅ Zhenhua Dong ⋅ Peng Zhao ⋅ Zhi-Hua Zhou
MEMGYM: a Long-Horizon Memory Environment for LLM Agents
Wujiang Xu ⋅ Yu Wang ⋅ Kai Mei ⋅ Kaiqu Liang ⋅ Zhenting Wang ⋅ Mingyu Jin ⋅ Han Zhang ⋅ Shi-Xiong Zhang ⋅ Wenyue Hua ⋅ Sambit Sahu ⋅ Dimitris N Metaxas
TraceNAS: Zero-shot Search for LLM Pruning via Gradient Trace Correlation
Prajna G. Malettira ⋅ Manish Nagaraj ⋅ Arjun Roy ⋅ Shubham Negi ⋅ Kaushik Roy
Knowledge‑Distilled Memory Editing for Plug‑and‑Play LLM Alignment
Haozheng Luo ⋅ Jiahao Yu ⋅ Wenxin Zhang ⋅ Jialong Li ⋅ Chenghao Qiu ⋅ Yimin Wang ⋅ Eric H Jiang ⋅ Ching-Yuen Huang ⋅ Chenwei Xu ⋅ Jerry Y Hu ⋅ Yan Chen ⋅ Binghui Wang ⋅ Xinyu Xing ⋅ Han Liu
Algorithmic Primitives of Theory of Mind Benchmarks
Qiutong Yi ⋅ Wang Bill Zhu ⋅ Robin Jia ⋅ Jesse Thomason
Inject a Turn, Not a Token: Learning Where to Redirect LLM Reasoning for Self-Correction
Jiwon Jeon ⋅ Myungsik Cho ⋅ Jeonghye Kim ⋅ Youngchul Sung
Test-Time Generation
Zefan Cai ⋅ Hao Tan ⋅ Yuchen Zhu ⋅ Tianyuan Zhang ⋅ Sai Bi ⋅ Haoyi Qiu ⋅ Yicong Hong ⋅ Jason Kuen ⋅ Haoliang Wang ⋅ Haozhe Zhao ⋅ Ruiyi Zhang ⋅ Ryan Rossi ⋅ Wanrong Zhu ⋅ Molei Tao ⋅ Wen Xiao ⋅ Rajiv Jain ⋅ Junjie Hu ⋅ Jiuxiang Gu
Clipping-Free Policy Optimization for Large Language Models
Ömer Çağatan ⋅ Baris Akgun ⋅ Gözde G Şahin ⋅ Xuandong Zhao
Multi-Head Attention Residuals
Cheng Luo ⋅ Zefan Cai ⋅ Junjie Hu
Gaze Attention: Selective Visual Attention for Multimodal Large Language Models
Junha Song ⋅ Byeongho Heo ⋅ Geonmo Gu ⋅ Jaegul Choo ⋅ Dongyoon Han ⋅ Sangdoo Yun
When Hidden States Drift: Can KV Caches Rescue Long-Range Speculative Decoding?
Tianyu Liu ⋅ Yuhao Shen ⋅ Xinyi Hu ⋅ Zhang Baolin ⋅ Hengxin Zhang ⋅ Jun Dai ⋅ Shuang Ge ⋅ Jun Zhang ⋅ Chen Lei ⋅ Yue Li ⋅ Mingcheng Wan ⋅ Xiaoyan Sun ⋅ Xiao Sun
Rethinking On-Policy Self-Distillation for Thinking Models
Simran Kaur ⋅ Narutatsu Ri ⋅ Yinghui He ⋅ Liam H Fowl ⋅ Sanjeev Arora
Attention Sinks and Outliers in Attention Residuals
Haoran Dai ⋅ Haozheng Luo ⋅ Ching-Yuen Huang ⋅ Shaoyang Zhang ⋅ Xi Chen ⋅ Eric H Jiang ⋅ Yijiang Li ⋅ Chenghao Qiu ⋅ Chenwei Xu ⋅ Zhenyu Pan ⋅ Jiakun Fan ⋅ Haotian Zhang ⋅ Dimitris N Metaxas ⋅ Binghui Wang ⋅ Yan Chen
Early Stopping for Large Reasoning Models via Confidence Dynamics
Parsa Hosseini ⋅ Sumit Nawathe ⋅ Mahdi Salmani ⋅ Meisam Razaviyayn ⋅ Soheil Feizi
Can RLVR Recover Reasoning Quality Missing from Mid-Training?
Tianyi Zhang ⋅ Stella Li ⋅ Yulia Tsvetkov
Know When to Stop: Segment-Level Credit Assignment for Reducing Overthinking
Chia-Hsuan Lee ⋅ Sihui Dai ⋅ Mingyang Zhou ⋅ Isha Slavin ⋅ Shi-Xiong Zhang ⋅ Sambit Sahu ⋅ William M Campbell
Playing with Fire: What Transfers When RL Trains a Language Agent?
Mahesh Ramesh ⋅ Kaousheik Jayakumar ⋅ Hemanth Ram ⋅ Pavan Thodima ⋅ Ramani Duraiswami ⋅ Dinesh Manocha ⋅ Aniket Rege ⋅ Emmanouil-Vasileios Vlatakis-Gkaragkounis
Last Layer Key-Value Cache is All You Need
Zefan Cai ⋅ Sai Bi ⋅ Lin Zhang ⋅ Yicong Hong ⋅ Chongjian GE ⋅ Xun Huang ⋅ Jiuxiang Gu ⋅ Junjie Hu ⋅ Hao Tan
Test-Time Training with Next-Token Prediction
Xuan Ouyang ⋅ Zefan Cai ⋅ Junjie Hu
Reason Wide, Not Deep: Amortizing the Reasoning Premium into Distilled Skills
Agamdeep Singh ⋅ Srishti Gautam ⋅ Priyanshu Gupta ⋅ Nikita Mehrotra ⋅ Tanmay Bakshi ⋅ Sumit Gulwani
Universal Test-Time Training
Zefan Cai ⋅ Qinzhe Hu ⋅ Ziqiao Ma ⋅ Hao Tan ⋅ Junjie Hu
SODA: Semi On-Policy Black-Box Distillation for Large Language Models
Xiwen Chen ⋅ Jingjing Wang ⋅ Wenhui Zhu ⋅ Peijie Qiu ⋅ XUANZHAO DONG ⋅ Yueyue Deng ⋅ Hejian Sang ⋅ Zhipeng Wang ⋅ Alborz Geramifard ⋅ Feng Luo
CRAFT: Reinforcement Learning from What Models Think, Not What They Say
Haozheng Luo ⋅ Yimin Wang ⋅ Haoran Dai ⋅ Jiahao Yu ⋅ Binghui Wang ⋅ Yan Chen