Arman Cohan

Papers on Lattice

Total citations

Topics

h-index

Research focus

Eval Frameworks & Benchmarks (2)Tool Use & Agents (2)Reasoning & Chain-of-Thought (1)RLHF & Preference Learning (1)Scientific Discovery & Drug Design (1)

Frequent co-authors

Xiao Zhou (1)Siyue Zhang (1)Jinbiao Wei (1)Tingyu Song (1)

Papers (3)

Jun 22, 2026

Xiao Zhou +5Jun 22, 2026·also University of Chineses Academy of Chineses

GUI vs. CLI: Execution Bottlenecks in Screen-Only and Skill-Mediated Computer-Use Agents

Skill augmentation can elevate CLI agent performance beyond GUI counterparts, revealing that the true challenge lies in skill coverage rather than model capability.

Xiao Zhou, Siyue Zhang, Jinbiao Wei +3

Eval Frameworks & Benchmarks Tool Use & Agents

Jun 17, 2026

Central Conservatory of MusicJun 17, 2026·also Yale

Rethinking Reward Supervision: Rubric-Conditioned Self-Distillation

Rubric-Conditioned Self-Distillation transforms vague scalar rewards into precise, actionable feedback, leading to significant improvements in reasoning model performance.

Siyi Gu, Jialin Chen, Sophia Zhou +2

Reasoning & Chain-of-Thought RLHF & Preference Learning

Jun 10, 2026

MIT CSAILJun 10, 2026·also BAIR, Microsoft Research, Stanford HAI, Corresponding authors. +8

Benchmarking AI Agents for Addressing Scientific Challenges Across Scales

Current AI agents excel in structured tasks but falter at generating novel insights and tackling open-ended scientific challenges.

Tianyu Liu, Allen Wang, Antonia Panescu +30

Eval Frameworks & Benchmarks Scientific Discovery & Drug Design Tool Use & Agents

Search

Arman Cohan

Research focus

Frequent co-authors

Papers (3)