“On the Promise and Limits of Training LLM Search Agents in Fictional Worlds” — by Anmol Kabra, Swathi Saravana Selvam, AI-MI PhD students Albert Gong, Chao Wan and Christian Belardi, Dongyoung Go, Katie Z. Luo and AI-MI Associate Director Kilian Weinberger (Cornell University) — was selected as an oral presentation at the COLM 2026 workshop “Learning from Situated and Embodied Interaction” (San Francisco, October 9, 2026). It is the workshop version of the AI-MI–supported preprint PhantomEnvironments: Training LLM Agents in Fictional Worlds. Congratulations to the team!
Source: COLM 2026 workshop: Learning from Situated and Embodied Interaction