Dropping some bonus content related to our “full-stack” excellence 😁
We recently presented at
@COLM_conf 2025 in Montréal—in collaboration with
@Princeton and
@UTAustin—advancing long-horizon reasoning, multi-agent coordination, and social strategy.
SPIN-Bench is a unified benchmark for Strategic Planning, Interaction, and Negotiation, spanning PDDL planning, competitive/co-op games, and Diplomacy-style bargaining.
🧵 Dive deeper to learn why we built SPIN-Bench and how it advances frontier AI reasoning.