影响力指数

27.99/100

前 26.9%

全站排名 #17,286

发表论文2 篇

平均评分7.3

年均产出2.0 篇/年

Maximilian Mozes

Researcher@Cohere·英国·OpenReview

研究方向

adversarial machine learning · natural language processing

Reverse Engineering Human Preferences with Reinforcement Learning

NeurIPS 2025Spotlight

Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models

ICLR 2025Poster

合作者 (13)

Dwaraknath Gnaneshwar

Edward Grefenstette

Siddhartha Rao Kamalakara