view article Article Field Notes: Why Muon "Hollows Out" in RL (and What We Plan To DO Next) bird-of-paradise β’ Feb 1 β’ 4
SVBench: Evaluation of Video Generation Models on Social Reasoning Paper β’ 2512.21507 β’ Published Dec 25, 2025 β’ 8
WildWorld: A Large-Scale Dataset for Dynamic World Modeling with Actions and Explicit State toward Generative ARPG Paper β’ 2603.23497 β’ Published Mar 24 β’ 92
Reasoning Models Struggle to Control their Chains of Thought Paper β’ 2603.05706 β’ Published Mar 5 β’ 39
PyVision-RL: Forging Open Agentic Vision Models via RL Paper β’ 2602.20739 β’ Published Feb 24 β’ 32
SVBench: Evaluation of Video Generation Models on Social Reasoning Paper β’ 2512.21507 β’ Published Dec 25, 2025 β’ 8 β’ 3
Running on Zero Agents Featured 982 MMAudio β generating synchronized audio from video/text π 982 Generate synchronized audio for videos or from text prompts
deepseek-ai/DeepSeek-V3-0324 Text Generation β’ 685B β’ Updated Mar 27, 2025 β’ 930k β’ β’ 3.15k
T3M: Text Guided 3D Human Motion Synthesis from Speech Paper β’ 2408.12885 β’ Published Aug 23, 2024 β’ 13
Running on Zero Agents Featured 1.02k IP-Adapter-FaceID π§ 1.02k Generate AI images featuring your own face