view article Article autotrust/JEV-27B: fast, calibrated decisions and full reasoning from one open model autotrust • 10 days ago • 218
VDiff-Bench: A Challenging Benchmark for Fine-Grained Image Difference Identification Paper • 2609.06245 • Published Sep 5 • 28
Scores Alone Do Not Prove Discovery: The Discovery Certification Protocol for Auditing AI Research Agents Paper • 2609.09219 • Published about 1 month ago • 16