影响力指数

32.85/100

前 21.3%

全站排名 #13,702

发表论文4 篇

平均评分4.3

年均产出2.0 篇/年

Aram H. Markosyan

Researcher@Facebook·美国·OpenReview

ProofOptimizer: Training Language Models to Simplify Proofs without Human Demonstrations

ICLR 2026Poster

With Great Backbones Comes Great Adversarial Transferability

ICLR 2026Withdrawn

Continual Learning via Sparse Memory Finetuning

ICLR 2026Desk Rejected

Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts

NeurIPS 2024Poster

合作者 (20)

Jack Parker-Holder

Jakob Nicolaus Foerster

Mikayel Samvelyan

Roberta Raileanu