PutnamBench: A Multilingual Competition-Mathematics Benchmark for Formal Theorem-Proving [paper]
George Tsoukalas · Jasper Lee · John Jennings · Jimmy Xin · Michelle Ding · Michael Jennings · Amitayush Thakur · Swarat Chaudhuri
🏆 Best Paper Award

Progress or Regress? Self-Improvement Reversal in Post-training [paper] 
Ting Wu · Xuefeng Li · Pengfei Liu
🏆 Honorable Mention Award

Learning to Reason by Failing: Offline RL on Sub-optimal Rollouts Scales Synthetic Data by 8x [paper]
Amrith Setlur · Saurabh Garg · Xinyang Geng · Naman Garg · Virginia Smith · Aviral Kumar

Lean4trace: Data augmentation for neural theorem proving in Lean [paper]
Vasilii Nesterov · Yermek Kapushev · Mikhail Burtsev 

Efficient Linear System Solver with Transformers [paper]
Max Vladymyrov · Johannes Von Oswald · Nolan Miller · Mark Sandler 

Large Language Models Can Self-Correct with Minimal Effort [paper]
Zhenyu Wu · Qingkai Zeng · Zhihan Zhang · Zhaoxuan Tan · Chao Shen · Meng Jiang 

Teaching Large Language Models to Reason with Reinforcement Learning [paper]
Alexander Havrilla · Yuqing Du · Sharath Chandra Raparthy · Christoforos Nalmpantis · Jane Dwivedi-Yu · Eric Hambro · Sainbayar Sukhbaatar · Roberta Raileanu

AI for an inverse problem: Physical model solving quantum gravity [paper]
Koji Hashimoto · Koshiro Matsuo · Masaki Murata · Gakuto Ogiwara · Daichi Takeda  

Specify What? A Case-Study using GPT-4 and Formal Methods For Specification Synthesis [paper]
George Granberry · Wolfgang Ahrendt · Moa Johansson 

DACO: Towards Application-Driven and Comprehensive Data Analysis via Code Generation [paper]
Xueqing Wu · Rui Zheng · Jingzhen Sha · Te-Lin Wu · Hanyu Zhou · Mohan Tang · Kai-Wei Chang · Nanyun Peng · Haoran Huang 

Distilling LLMs’ Decomposition Abilities into Compact Language Models [paper]
Denis Tarasov · Kumar Shridhar 

Progressive-Hint Prompting Improves Reasoning in Large Language Models [paper]
Chuanyang Zheng · Zhengying Liu · Enze Xie · Zhenguo Li · Yu Li 

VerityMath: Advancing Mathematical Reasoning by Self-Verification Through Unit Consistency [paper]
Vernon Yan Han Toh · Ratish Puduppully · Nancy Chen

Smart Vision-Language Reasoners [paper]
Denisa Roberts · Lucas Roberts 

Pre-Calc: Learning to Use the Calculator Improves Numeracy in Language Models [paper]
Vishruth Veerendranath · Vishwa Shah · Kshitish Ghate

Learning Efficient Recursive Numeral Systems via Reinforcement Learning [paper]
Jonathan Thomas · Andrea Silvi · Devdatt Dubhashi · Emil Carlsson · Moa Johansson 

More Details, Please: Improving Autoformalization with More Detailed Proofs [paper]
Guillem Tarrach · Albert Jiang · Daniel Raggi · Wenda Li · Mateja Jamnik 

Technical Report for ICML 2024 Automated Math Reasoning Challenge: Solving Optimization Problems with Open Source Large Language Model [paper]
Duc M. Nguyen · Sungahn Ko 

Advancing LLM Reasoning Generalists with Preference Trees [paper]
Lifan Yuan · Ganqu Cui · Hanbin Wang · Ning Ding · Xingyao Wang · Jia Deng · Boji Shan · Huimin Chen · Ruobing Xie · Yankai Lin · Zhenghao Liu · Bowen Zhou · Hao Peng · Zhiyuan Liu · Maosong Sun

GeomVerse: A Systematic Evaluation of Large Models for Geometric Reasoning [paper]
Mehran Kazemi · Hamidreza Alvari · Ankit Anand · Jialin Wu · Xi Chen · Radu Soricut 

Metacognitive Capabilities of LLMs: An Exploration in Mathematical Problem Solving  [paper]
Aniket Didolkar · Anirudh Goyal · Rosemary Nan Ke · Siyuan Guo · Michal Valko · Timothy Lillicrap · Danilo J. Rezende · Yoshua Bengio · Michael Mozer · Sanjeev Arora