Why Does ChatGPT “Delve” So Much? Exploring the Sources of Lexical Overrepresentation in Large Language…
Hybrid Preference Optimization for Alignment: Provably Faster Convergence Rates by Combining Offline Preferences with Online…
Solving the Inverse Alignment Problem for Efficient RLHFby Shambhavi Krishna, Aishwarya SahooFirst submitted to arxiv…
Does RLHF Scale? Exploring the Impacts From Data, Model, and Methodby Zhenyu Hou, Pengfan Du,…
Self-Improvement in Language Models: The Sharpening Mechanismby Audrey Huang, Adam Block, Dylan J. Foster, Dhruv…
Yi-Lightning Technical Reportby Alan Wake, Bei Chen, C.X. Lv, Chao Li, Chengen Huang, Chenglin Cai,…
Self-Generated Critiques Boost Reward Modeling for Language Modelsby Yue Yu, Zhengxing Chen, Aston Zhang, Liang…
Multimodal large language model for wheat breeding: a new exploration of smart breedingby Guofeng Yang,…
Continual SFT Matches Multimodal RLHF with Negative Supervisionby Ke Zhu, Yu Wang, Yanpeng Sun, Qiang…
Value Imprint: A Technique for Auditing the Human Values Embedded in RLHF Datasetsby Ike Obi,…