Self-Generated Feedback Destabilizes Test-Time Training: A Causal Decomposition of Long-Horizon Adaptation Paper • 2610.05076 • Published 3 days ago • 18
DiSCO: Defending text-to-image generation through distribution-guided contrastive prompt optimization Paper • 2608.17067 • Published Aug 17 • 22
LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know? Paper • 2605.28721 • Published May 27 • 15