1TournO: Tournament Optimization for Non-Verifiable RL (opens in new tab)(github.com)3leonardtang2d ago0
2j1-micro and j1-nano: Tiny (0.6B, 1.7B) and Mighty Reward Models (opens in new tab)(github.com)3leonardtang10mo ago0
3Verdict: A Library for Scaling Judge-Time Compute (opens in new tab)(twitter.com)3leonardtang1y ago0
6Cascade: A fast, automated, multi-turn LLM jailbreaking method (opens in new tab)(twitter.com)2leonardtang1y ago0
11Sphynx: Fuzz Testing Hallucination Detection Models (opens in new tab)(github.com)2leonardtang1y ago0
13Thorn in a HaizeStack test for evaluating long-context adversarial robustness (opens in new tab)(github.com)19leonardtang1y ago11
14Thorn in a HaizeStack Long-Context Jailbreak Test (opens in new tab)(github.com)5leonardtang1y ago0
