STEPQuant: When and Where Errors Matter in Delta-Rule Recurrent State Quantization Paper • 2609.38169 • Published Sep 29 • 37
MaLiang-Harness: A Programmable Path to Image and Video Generation Paper • 2609.34309 • Published Sep 28 • 89
DecepEval: A Benchmark for Evaluating Deception in LLM Agents Paper • 2610.07967 • Published Oct 6 • 37