Enabling two settings tripled our scores on the ARC-AGI-3 benchmark
tedsanders
23 points
0 comments
July 29, 2026
Related Discussions
Found 5 related stories in 372.3ms across 15,380 title embeddings via pgvector HNSW
- From 0% to 36% on Day 1 of ARC-AGI-3 lairv · 60 pts · March 27, 2026 · 60% similar
- ARC-AGI-3 lairv · 328 pts · March 25, 2026 · 52% similar
- OpenRCA benchmark – Improving Claude's root cause analysis accuracy by 12 pp behat · 11 pts · March 11, 2026 · 51% similar
- ARC-AGI Leaderboard rzk · 168 pts · July 25, 2026 · 48% similar
- AlphaEvolve: Gemini-powered coding agent scaling impact across fields berlianta · 274 pts · May 07, 2026 · 48% similar