ARC-AGI-3 scaffolding, not new models, nearly doubled community leaderboard scores to 55.89% in four days after Tufa Labs' Duck harness went open-source. OpenAI's controlled experiment showed two ...
Hosted on MSN
ARC-AGI-3 gets open-source agent that writes Python world models instead of neural weights
A German academic research group today posted an open-source AI agent to arXiv that approaches ARC-AGI-3 — the interactive benchmark where frontier models including Claude Opus 5 still cannot reliably ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results