Edan Toledo

AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents

arXiv 2026

Investigating Memory in Model-Free RL with POPGym Arcade

arXiv 2025

The Automated LLM Speedrunning Benchmark: Reproducing NanoGPT Improvements

arXiv 2025

Abhishek Charnalia

Alisia Lupidi

Derek Dunfield

Despoina Magka

Jakob Foerster

Jean-Christophe Gagnon-Audet

Karen Hambardzumyan

Kelvin Niu

Lucia Cipolina-Kun

Martin Josifoski