@SpaceXAI On ARC-AGI-3, Grok 4.7 scores 1.8% (vs Grok 4.7's 2.1%) in the standard harness, which lets models carry forward notes between turns, and 10.0% in a new provider adapter harness, which…
Original title@SpaceXAI On ARC-AGI-3, Grok 4.7 scores 1.8% (vs Grok 4.7's 2.1%) in the standard harness, which lets models carry forward notes between ...
AISummary
…preserves opaque reasoning and enables auto compaction.
Source: ARC Prize · x.com