16x Faster On-Device Video Generation: Qualcomm's ReHyAt Distills Attention in 160 GPU HoursJanuary 2026VisionLearn more
VLM Hallucinations Exposed: VIB-Probe Pinpoints and Suppresses Faulty Attention HeadsJanuary 2026SafetyLearn more
Training-Free Fix Boosts Vision-Language Models 3 Points by Correcting Attention ErrorsJanuary 2026VisionLearn more
Gold Medal at IMO and IOI: DeepSeek-V3.2 Matches GPT-5 with Open WeightsJanuary 2026TrainingLearn more
Why Chain-of-Thought Works: Researchers Find a Single 'Reasoning Switch' in LLMsJanuary 2026LLMsLearn more
Zero Training Data, Full Performance: Dr. Zero Matches Supervised Search AgentsJanuary 2025TrainingLearn more
6% Better Math Reasoning in Fewer Tokens: Multiplex Thinking Merges Multiple Paths into OneJanuary 2025TrainingLearn more
40% Faster Video from Single Images: Pixel-to-4D Predicts Dynamic 3D Gaussians in One PassJanuary 2025VisionLearn more