Daily Papers of 2026-01-05

1. NeoVerse: Enhancing 4D World Model with in-the-wild Monocular Videos 111 upvotes, #1 of 2026-01-05
2. Youtu-Agent: Scaling Agent Productivity with Automated Generation and Hybrid Policy Optimization 108 upvotes, #2 of 2026-01-05
3. Taming Hallucinations: Boosting MLLMs' Video Understanding via Counterfactual Video Generation 57 upvotes, #3 of 2026-01-05
4. Avatar Forcing: Real-Time Interactive Head Avatar Generation for Natural Conversation 51 upvotes, #4 of 2026-01-05
5. Nested Learning: The Illusion of Deep Learning Architectures 34 upvotes, #5 of 2026-01-05
6. SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learning 33 upvotes, #6 of 2026-01-05
7. AdaGaR: Adaptive Gabor Representation for Dynamic Scene Reconstruction 30 upvotes, #7 of 2026-01-05
8. Deep Delta Learning 29 upvotes, #8 of 2026-01-05
9. The Reasoning-Creativity Trade-off: Toward Creativity-Driven Problem Solving 18 upvotes, #9 of 2026-01-05
10. Taming Preference Mode Collapse via Directional Decoupling Alignment in Diffusion Reinforcement Learning 12 upvotes, #10 of 2026-01-05
11. Diversity or Precision? A Deep Dive into Next Token Prediction 7 upvotes, #11 of 2026-01-05
12. MorphAny3D: Unleashing the Power of Structured Latent in 3D Morphing 5 upvotes, #12 of 2026-01-05
13. Fast-weight Product Key Memory 4 upvotes, #13 of 2026-01-05
14. InfoSynth: Information-Guided Benchmark Synthesis for LLMs 2 upvotes, #14 of 2026-01-05
15. Can We Trust AI Explanations? Evidence of Systematic Underreporting in Chain-of-Thought Reasoning 2 upvotes, #14 of 2026-01-05

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.