Stanford AI Lab Papers and Talks at ICLR 2026

April 22, 2026

The International Conference on Learning Representations (ICLR) 2026 is being hosted in Rio de Janeiro from April 23rd – April 27th. We’re excited to share all the work from SAIL being presented, with links to papers, websites, blog posts, and videos below. Feel free to reach out to the contact authors directly to learn more about the work happening at Stanford!

List of Accepted Papers

AbdCTBench: Learning Clinical Biomarker Representations from Abdominal Surface Geometry

Authors: Muhammad Ahmed Chaudhry, Suhana Bedi, Pola Lydia Lagari, Brian T Layden, William Galanter, Ayis Pyrros, Sanmi Koyejo
Contact: mahmedch@stanford.edu
Links: Paper | Website
Keywords: computer vision for healthcare, radiology, computed tomography (CT), vision transformers, CNNs

AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimization

Authors: Genghan Zhang, Shaowei Zhu, Anjiang Wei, Zhenyu Song, Allen Nie, Zhen Jia, Nandita Vijaykumar, Yida Wang, Kunle Olukotun
Contact: zgh23@stanford.edu
Award nominations: ICLR 2026 MALGAI workshop (Oral)
Links: Paper | Website | Blog Post
Keywords: self-improving LLM agents, kernel optimization, AI accelerator

Addressing Divergent Representations from Causal Interventions on Neural Networks

Authors: Satchel Grant, Simon Jerome Han, Alexa R. Tartaglini, Christopher Potts
Contact: grantsrb@stanford.edu
Award nominations: ICLR 2026 Oral Presentation
Links: Paper | Website
Keywords: causal interventions, computational pathways, DAS, patching

An Information Theoretic Perspective on Agentic System Design

Authors: Shizhe He, Avanika Narayan, Ishan S. Khare, Scott W. Linderman, Christopher Ré, Dan Biderman
Contact: shizhehe@stanford.edu
Links: Paper | Blog Post
Keywords: information bottleneck, rate-distortion theory, agentic collaboration, large language models, scaling laws

ATLAS: Adaptive Transfer Scaling Laws for Multilingual Pretraining, Finetuning, and Decoding the Curse of Multilinguality

Authors: Shayne Longpre, Sneha Kudugunta, Niklas Muennighoff, I Hsu, Isaac Caswell, Alex Pentland, Sercan Arik, Chen-Yu Lee, Sayna Ebrahimi
Contact: slongpre@media.mit.edu
Links: Paper
Keywords: scaling laws, large language models

AutoMetrics: Approximate Human Judgements with Automatically Generated Evaluators

Authors: Michael J. Ryan, Yanzhe Zhang, Amol Salunkhe, Yi Chu, Di Xu, Diyi Yang
Contact: mryan0@stanford.edu
Links: Paper | Website
Keywords: evaluation, metrics, data-constrained, human-centered, LLM-as-a-judge

Blending Training and Deployment Data with Weighted Expert Ensembles for Post-hoc LLM Calibration

Authors: Aishwarya Mandyam, Wenhui Sophia Lu, Wing Hung Wong, John Duchi, Barbara E. Engelhardt
Contact: am2@stanford.edu
Links: Paper
Keywords: calibration, distribution shift

Can Large Language Models Match the Conclusions of Systematic Reviews?

Authors: Christopher Polzak, Alejandro Lozano, Min Woo Sun, James Burgess, Yuhui Zhang, Kevin Wu, Chia-Chun Chiang, Jeffrey J Nirschl, Serena Yeung-Levy
Contact: clcp@cs.stanford.edu
Links: Paper | Website
Keywords: benchmarks, multi-document reasoning, medical AI

Causal Interpretation of Neural Network Computations with Contribution Decomposition

Authors: Joshua Brendan Melander, Zaki Alaoui, Shenghua Liu, Surya Ganguli, Stephen A. Baccus
Contact: sliu24@stanford.edu
Links: Paper
Keywords: mechanistic interpretability, neuroscience, XAI, AI safety, tool development

Collaborative Gym: A Framework for Enabling and Evaluating Human-Agent Collaboration

Authors: Yijia Shao, Vinay Samuel, Yucheng Jiang, John Yang, Diyi Yang
Contact: shaoyj@stanford.edu
Links: Paper | Website | Video
Keywords: language model agents, human-AI collaboration, human-in-the-loop, evaluation

Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning

Authors: Moo Jin Kim, Yihuai Gao, Tsung-Yi Lin, Yen-Chen Lin, Yunhao Ge, Grace Lam, Percy Liang, Shuran Song, Ming-Yu Liu, Chelsea Finn, Jinwei Gu
Contact: moojink@stanford.edu
Links: Paper | Website | Video
Keywords: world models, robotics, manipulation, model-based planning, imitation learning, video generation

Cost-of-Pass: An Economic Framework for Evaluating Language Models

Authors: Mehmet Hamza Erol, Batu El, Mirac Suzgun, Mert Yuksekgonul, James Y Zou
Contact: mhamza@stanford.edu
Links: Paper | Website
Keywords: economic evaluation framework, language-model evaluation, cost-performance trade-off, inference time techniques

CTC-DRO: Robust Optimization for Reducing Language Disparities in Speech Recognition

Authors: Martijn Bartelds, Ananjan Nandi, Moussa Koulako Bala Doumbouya, Dan Jurafsky, Tatsunori Hashimoto, Karen Livescu
Contact: bartelds@stanford.edu
Links: Paper | Website | Video
Keywords: distributionally robust optimization, deep learning, robustness, speech recognition

DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search

Authors: Fang Wu, Weihao Xuan, Heli Qi, Ximing Lu, Aaron Tu, Li Erran Li, Yejin Choi
Contact: fangwu97@stanford.edu
Links: Paper | Website
Keywords: MCTS, RLVR

Discrete Diffusion Trajectory Alignment via Stepwise Decomposition

Authors: Jiaqi Han, Austin Wang, Minkai Xu, Wenda Chu, Meihua Dang, Haotian Ye, Huayu Chen, Yisong Yue, Stefano Ermon
Contact: jiaqihan@stanford.edu
Keywords: discrete diffusion, language models, reinforcement learning

Distributional Machine Unlearning via Selective Data Removal

Authors: Youssef Allouah, Rachid Guerraoui, Sanmi Koyejo
Contact: yallouah@stanford.edu
Links: Paper
Keywords: unlearning, theory, privacy, sample complexity, machine learning, statistical learning

DSL-Monkeys: Self-Generated In-Context Examples for Low-Resource GPU DSL Kernels

Authors: Nathan Paek*, Simon Guo*, Vishnu Sarukkai*, Willy Chan*, William Hu, Ethan Boneh, Simran Arora, Ludwig Schmidt, Kayvon Fatahalian, Azalia Mirhoseini
Contact: simonguo@stanford.edu
Links: Paper
Keywords: GPU, in-context learning, low-resource data

ENACT: Evaluating Embodied Cognition with World Modeling of Egocentric Interaction

Authors: Qineng Wang*, Wenlong Huang*, Yu Zhou, Hang Yin, Tianwei Bao, Jianwen Lyu, Weiyu Liu, Ruohan Zhang, Jiajun Wu, Li Fei-Fei, Manling Li (*Equal Contribution)
Contact: wenlongh@stanford.edu
Award nominations: Oral Presentation at ICLR 2026 Workshop on World Models, Outstanding Paper Award at ICLR 2026 Workshop on Lifelong Agents
Links: Paper | Website | Video
Keywords: embodied agents, vision language models, benchmarking, world modeling

From Tokens to Thoughts: How LLMs and Humans Trade Compression for Meaning

Authors: Chen Shani, Liron Soffer, Dan Jurafsky, Yann LeCun, Ravid Shwartz-Ziv
Contact: cshani@stanford.edu
Links: Paper
Keywords: human-centered NLP, machine cognition, information theory for LLMs

Front-Loading Reasoning: The Synergy between Pretraining and Post-Training Data

Authors: Syeda Nahida Akter, Shrimai Prabhumoye, Eric Nyberg, Mostofa Patwary, Mohammad Shoeybi, Yejin Choi, Bryan Catanzaro
Contact: yejinc@cs.stanford.edu
Links: Paper
Keywords: pretraining, supervised finetuning, reasoning, LLM

Geometry-aware 4D Video Generation for Robot Manipulation

Authors: Zeyi Liu, Shuang Li, Eric Cousineau, Siyuan Feng, Benjamin Burchfiel, Shuran Song
Contact: liuzeyi@stanford.edu
Links: Paper | Website | Video
Keywords: video generation, robot manipulation, 3D perception

Goedel-Prover-V2: Scaling Formal Theorem Proving with Scaffolded Data Synthesis and Self-Correction

Authors: Yong Lin, Shange Tang, Bohan Lyu, Ziran Yang, Jui-Hui Chung, Haoyu Zhao, Lai Jiang, Yihan Geng, Jiawei Ge, Jingruo Sun, Jiayun Wu, Jiri Gesi, Ximing Lu, David Acuna, Kaiyu Yang, Hongzhou Lin, Yejin Choi, Danqi Chen, Sanjeev Arora, Chi Jin
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: theorem proving, reasoning

Humanline: Online Alignment as Perceptual Loss

Authors: Sijia Liu, Niklas Muennighoff, Kawin Ethayarajh
Contact: kawin@uchicago.edu
Links: Paper
Keywords: humanline, KTO

HUME: Measuring the Human-Model Performance Gap in Text Embedding Tasks

Authors: Adnan El Assadi, Isaac Chung, Roman Solomatin, Niklas Muennighoff, Kenneth Enevoldsen
Contact: niklasm@stanford.edu
Links: Paper | Website
Keywords: embeddings, MTEB

In-the-Flow Agentic System Optimization for Effective Planning and Tool Use

Authors: Zhuofeng Li, Haoxiang Zhang, Seungju Han, Sheng Liu, Jianwen Xie, Yu Zhang, Yejin Choi, James Zou, Pan Lu
Contact: lupantech@gmail.com
Award nominations: ICLR Oral Presentation, Best Paper Runner-up, NeurIPS 2025 ER Workshop
Links: Paper | Website | Video
Keywords: LLM agents, reinforcement learning, tool use, multi-turn interaction

Kevin: Multi-Turn RL for Generating CUDA Kernels

Authors: Carlo Baronio*, Pietro Marsella*, Ben Pan*, Simon Guo, Silas Alberti
Contact: simonguo@stanford.edu
Links: Paper | Website | Blog Post
Keywords: multi-turn, RL, GPU kernel, code generation

Learning to Summarize User Information for Personalized Reinforcement Learning from Human Feedback

Authors: Hyunji (Alex) Nam, Yanming Wan, Mickel Liu, Peter F. Ahnn, Jianxun Lian, Natasha Jaques
Contact: hjnam@stanford.edu
Links: Paper | Website
Keywords: pluralistic alignment, LLM personalization, reward modeling for RLHF

LitmusValues: Will AI Tell Lies to Save Sick Children? Litmus-Testing AI Values Prioritization with AIRiskDilemmas

Authors: Yu Ying Chiu, Zhilin Wang, Sharan Maiya, Yejin Choi, Kyle Fish, Sydney Levine, Evan J Hubinger
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: AI values, value alignment, AI risk, dilemma

Markovian Transformers for Informative Language Modeling

Authors: Scott W Viteri, Max Lamparth, Peter Chatain, Clark Barrett
Contact: scottviteri@gmail.com
Links: Paper
Keywords: Markovian transformers, chain-of-thought reasoning, language model interpretability, causal reasoning, reinforcement learning, next-token prediction, GSM8K, large language models

Measuring LLM Novelty As The Frontier Of Original And High-Quality Output

Authors: Vishakh Padmakumar, Chen Yueh-Han, Jane Pan, Valerie Chen, He He
Contact: vishakhp@stanford.edu
Links: Paper | Blog Post
Keywords: generation, evaluation, memorization, novelty, benchmark, creativity

Mixture of Contexts for Long Video Generation

Authors: Shengqu Cai, Ceyuan Yang, Lvmin Zhang, Yuwei Guo, Junfei Xiao, Ziyan Yang, Yinghao Xu, Zhenheng Yang, Alan Yuille, Leonidas Guibas, Maneesh Agrawala, Lu Jiang, Gordon Wetzstein
Contact: shengqu@stanford.edu
Links: Paper | Website | Blog Post | Video
Keywords: video generation, generative models, sparse attention, movie generation

MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes

Authors: Yu Ying Chiu, Michael S. Lee, Rachel Calcott, Brandon Handoko, Paul de Font-Reaulx, Paula Rodriguez, Chen Bo Calvin Zhang, Ziwen Han, Udari Madhushani Sehwag, Yash Maurya, Christina Q Knight, Harry R. Lloyd, Florence Bacus, Mantas Mazeika, Bing Liu, Yejin Choi, Mitchell L Gordon, Sydney Levine
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: moral reasoning, reasoning evaluation, AI safety

Moving Beyond Medical Exams: A Clinician-Annotated Fairness Dataset of Real-World Tasks and Ambiguity in Mental Healthcare

Authors: Max Lamparth, Declan Grabb, Amy Franks, Scott Gershan, Kaitlyn N Kunstman, Aaron Lulla, Monika Drummond Roots, Manu Sharma, Aryan Shrivastava, Nina Vasan, Colleen Waickman
Contact: lamparth@stanford.edu
Links: Paper | Blog Post
Keywords: AI for healthcare, mental health, fairness, bias, dataset, language models, decision-making, uncertainty, expert annotation

mR3: Multilingual Rubric-Agnostic Reward Reasoning Models

Authors: David Anugraha, Shou-Yi Hung, Zilu Tang, Annie En-Shiun Lee, Derry Tanti Wijaya, Genta Indra Winata
Contact: davidanu@stanford.edu
Links: Paper | Website
Keywords: multilingual, reasoning, reward model, rubrics

Multiplayer Nash Preference Optimization

Authors: Fang Wu, Xu Huang, Weihao Xuan, Zhiwei Zhang, Yijia Xiao, Guancheng Wan, Xiaomin Li, Bing Hu, Peng Xia, Jure Leskovec, Yejin Choi
Contact: fangwu97@stanford.edu
Award nominations: Oral
Links: Paper | Website
Keywords: preference optimization, RLHF, LLM post-training

No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms

Authors: Joshua Kazdan, Abhay Puri, Rylan Schaeffer, Lisa Yu, Chris Cundy, Jason Stanley, Sanmi Koyejo, Krishnamurthy Dj Dvijotham
Contact: jkazdan@stanford.edu
Links: Paper
Keywords: jailbreaking, attacks, AI safety, red-teaming, fine-tuning, fine-tuning attacks

ODESteer: A Unified ODE-Based Steering Framework for LLM Alignment

Authors: Hongjue Zhao, Haosen Sun, Jiangtao Kong, Xiaochang Li, Qineng Wang, Liwei Jiang, Qi Zhu, Tarek F. Abdelzaher, Yejin Choi, Manling Li, Huajie Shao
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: LLM alignment, representation engineering, activation steering, ODE-based framework, barrier functions

OpenThoughts: Data Recipes for Reasoning Models

Authors: Etash Kumar Guha, Ryan Marten, Sedrick Keh, Negin Raoof, Georgios Smyrnis, Hritik Bansal, Marianna Nezhurina, Jean Mercat, Trung Vu, Zayne Rea Sprague, Ashima Suvarna, Benjamin Feuer, Leon Liangyu Chen, Zaid Khan, Eric Frankel, Sachin Grover, Caroline Choi, Niklas Muennighoff, Shiye Su, Wanjia Zhao, et al.
Contact: yejinc@cs.stanford.edu
Links: Paper
Keywords: reasoning, data, LLM

Optimal Aggregation Mechanisms for AI Benchmarking and Platinum Benchmarks

Authors: Andreas Haupt, Anka Reuel, Mykel Kochenderfer, Sanmi Koyejo
Contact: h4upt@stanford.edu
Links: Paper
Keywords: mechanism design, AI evaluation, benchmarking, multi-tasking

pi-Flow: Policy-Based Few-Step Generation via Imitation Distillation

Authors: Hansheng Chen, Kai Zhang, Hao Tan, Leonidas Guibas, Gordon Wetzstein, Sai Bi
Contact: hshchen@stanford.edu
Links: Paper | Website
Keywords: diffusion models, flow models, few-step generation, distillation

Polychromic Objectives for Reinforcement Learning

Authors: Jubayer Ibn Hamid, Ifdita Hasan Orney, Ellen Xu, Chelsea Finn, Dorsa Sadigh
Contact: jubayer@stanford.edu
Links: Paper
Keywords: reinforcement learning, exploration

PoseX: AI Defeats Physics-based Methods on Protein Ligand Cross-Docking

Authors: Yize Jiang, Xinze Li, Yuanyuan Zhang, Jin Han, Youjun Xu, Ayush Pandit, Zaixi Zhang, Mengdi Wang, Mengyang Wang, Chong Liu, Guang Yang, Yejin Choi, Yingzhou Lu, Wu-Jun Li, Tianfan Fu, Fang Wu, Junhong Liu
Contact: fangwu97@stanford.edu
Links: Paper | Website
Keywords: AI docking, AI co-folding, protein-ligand interaction, cross docking

Pre-training under infinite compute

Authors: Konwoo Kim, Suhas Kotha, Percy Liang, Tatsunori Hashimoto
Contact: konwoo@stanford.edu, kotha@stanford.edu
Award nominations: Oral
Links: Paper
Keywords: pre-training, data efficiency, scaling laws

Pretraining Scaling Laws for Generative Evaluations of Language Models

Authors: Rylan Schaeffer, Noam Itzhak Levi, Brando Miranda, Sanmi Koyejo
Contact: rschaef@cs.stanford.edu
Links: Paper
Keywords: language models, large language models, scaling laws, evaluations, generative evaluations, sampling

ProfBench: Multi-Domain Rubrics requiring Professional Knowledge to Answer and Judge

Authors: Zhilin Wang, Jaehun Jung, Ximing Lu, Shizhe Diao, Ellie Evans, Jiaqi Zeng, Pavlo Molchanov, Yejin Choi, Jan Kautz, Yi Dong
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: expert-annotated, professional knowledge, LLM judge, rubric evaluation

Relational Transformer: Toward Zero-Shot Foundation Models for Relational Data

Authors: Rishabh Ranjan, Valter Hudovernik, Mark Znidar, Charilaos Kanatsoulis, Roshan Upendra, Mahmoud Mohammadi, Joe Meyer, Tom Palczewski, Carlos Guestrin, Jure Leskovec
Contact: ranjanr@stanford.edu
Award nominations: Oral at AI for Tabular Data Workshop, NeurIPS 2025
Links: Paper | Website
Keywords: foundation models, tabular data, relational databases, transformer

RLP: Reinforcement as a Pretraining Objective

Authors: Ali Hatamizadeh, Syeda Nahida Akter, Shrimai Prabhumoye, Jan Kautz, Mostofa Patwary, Mohammad Shoeybi, Bryan Catanzaro, Yejin Choi
Contact: yejinc@cs.stanford.edu
Links: Paper
Keywords: reinforcement learning, pretraining, reasoning, large language models

SCRIBES: Web-Scale Script-Based Semi-Structured Data Extraction with Reinforcement Learning

Authors: Shicheng Liu, Kai Sun, Lisheng Fu, Xilun Chen, Xinyuan Zhang, Zhaojiang Lin, Rulin Shao, Yue Liu, Anuj Kumar, Wen-tau Yih, Xin Luna Dong
Contact: shicheng@cs.stanford.edu
Links: Paper
Keywords: semi-structured data, reinforcement learning, information extraction

SimBench: Benchmarking the Ability of Large Language Models to Simulate Human Behaviors

Authors: Tiancheng Hu, Joachim Baumann, Lorenzo Lupo, Nigel Collier, Dirk Hovy, Paul Röttger
Contact: th656@cam.ac.uk
Links: Paper | Website
Keywords: human behavior simulation, large language models, benchmarking, computational social science, human-AI alignment, calibration, human-centered AI

SimpleToM: Exposing the Gap between Explicit ToM Inference and Implicit ToM Application in LLMs

Authors: Yuling Gu, Oyvind Tafjord, Hyunwoo Kim, Jared Moore, Ronan Le Bras, Peter Clark, Yejin Choi
Contact: yejinc@cs.stanford.edu
Links: Paper
Keywords: theory of mind, social reasoning, LLM benchmark, mental state, behavior, judgment, false belief

Spatial Mental Modeling from Limited Views

Authors: Qineng Wang, Baiqiao Yin, Pingyue Zhang, Jianshu Zhang, Kangrui Wang, Zihan Wang, Jieyu Zhang, Keshigeyan Chandrasegaran, Han Liu, Ranjay Krishna, Saining Xie, Jiajun Wu, Li Fei-Fei, Manling Li
Contact: qinengwang2029@u.northwestern.edu
Award nominations: Best Paper Honorable Mention at NeurIPS 2025 LAW Workshop, Best Paper Award at ICCV 2025 SP4V Workshop, The Best of ICCV featured by Voxel51
Links: Paper | Website
Keywords: vision language models, VLMs, multimodal language models, spatial intelligence, spatial reasoning

Spectrum Tuning: Post-Training for Distributional Coverage and In-Context Steerability

Authors: Taylor Sorensen, Benjamin Newman, Jared Moore, Chan Young Park, Jillian Fisher, Niloofar Mireshghallah, Liwei Jiang, Yejin Choi
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: post-training, language models, distributional learning, alignment, pluralistic alignment, uncertainty estimation

Speculative Speculative Decoding

Authors: Tanishq Kumar, Tri Dao, Avner May
Contact: tanishq@stanford.edu
Links: Paper
Keywords: inference, LLMs, systems

Theory of Space: Can Foundation Models Construct Spatial Beliefs through Active Exploration?

Authors: Pingyue Zhang, Zihan Huang, Yue Wang, Jieyu Zhang, Letian Xue, Zihan Wang, Qineng Wang, Keshigeyan Chandrasegaran, Ruohan Zhang, Yejin Choi, Ranjay Krishna, Jiajun Wu, Li Fei-Fei, Manling Li
Contact: pingyuezhang2029@u.northwestern.edu
Links: Paper | Website | Blog Post
Keywords: large language model, vision-language model, spatial reasoning, spatial agent, active exploration

ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning

Authors: Jiawei Gu, Yunzhuo Hao, Huichen Will Wang, Linjie Li, Michael Qizhe Shieh, Yejin Choi, Ranjay Krishna, Yu Cheng
Contact: yejinc@cs.stanford.edu
Links: Paper | Website
Keywords: multimodal reasoning, interleaved chain-of-thought, unified model

TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models

Authors: Yue Huang, Chujie Gao, Siyuan Wu, Haoran Wang, Xiangqi Wang, Jiayi Ye, Yujun Zhou, Yanbo Wang, Jiawen Shi, Qihui Zhang, Han Bao, Zhaoyi Liu, Yuan Li, Tianrui Guan, Peiran Wang, Haomin Zhuang, Dongping Chen, Kehan Guo, Andy Zou, Bryan Hooi, et al.
Contact: yhuang37@nd.edu
Links: Paper | Website
Keywords: trustworthiness, generative model, large language model, vision-language model, dynamic evaluation, benchmark

Keep on top of the latest SAIL Blog posts via , , or email: