RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversations Paper • 2610.01780 • Published Oct 1 • 40
4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes Paper • 2610.03715 • Published Oct 2 • 22
Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression Paper • 2609.36322 • Published Sep 28 • 40
EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments? Paper • 2609.37686 • Published Sep 29 • 55