ASCENT: Online Test-Time Training of Long-Horizon Agents via Self-Distillation of Verified ExperienceResearchOct 6, 2026
What Does a Harness Repair? A Preregistered Study of Visibility, Baseline Adequacy and Evaluation DefectsResearchOct 6, 2026
AdaSpark: Adaptive DSpark with Online Learning for Tree Verification and N-gram FillResearchOct 6, 2026
CurveCodec 2: Skeleton-agnostic animation compression with a learned entropy modelResearchOct 6, 2026
Erased, Rerouted, or Rescaled? Post-Training and the Causal Quotient of a Language Model's Belief StateResearchOct 6, 2026