TournO: Tournament Optimization for Non-Verifiable RLgithub.com3 points·by leonardtang·قبل 4 أشهر·0 comments
j1-micro and j1-nano: Tiny (0.6B, 1.7B) and Mighty Reward Modelsgithub.com3 points·by leonardtang·السنة الماضية·0 comments
Verdict: A Library for Scaling Judge-Time Computetwitter.com3 points·by leonardtang·السنة الماضية·0 comments
Cascade: A fast, automated, multi-turn LLM jailbreaking methodtwitter.com2 points·by leonardtang·قبل سنتين·0 comments
Sphynx: Fuzz Testing Hallucination Detection Modelsgithub.com2 points·by leonardtang·قبل سنتين·0 comments
Thorn in a HaizeStack test for evaluating long-context adversarial robustnessgithub.com19 points·by leonardtang·قبل سنتين·11 comments
Thorn in a HaizeStack Long-Context Jailbreak Testgithub.com5 points·by leonardtang·قبل سنتين·0 comments
leonardtang·قبل 3 سنوات·discussPrecisely -- watermarks are an obvious example of this. To me, this is THE path forward for AI content detection.