Daily Papers of 2026-04-06

1. GrandCode: Achieving Grandmaster Level in Competitive Programming via Agentic Reinforcement Learning 351 upvotes, #1 of 2026-04-06
2. InCoder-32B-Thinking: Industrial Code World Model for Thinking 228 upvotes, #2 of 2026-04-06
3. Self-Distilled RLVR 157 upvotes, #3 of 2026-04-06
4. A Simple Baseline for Streaming Video Understanding 72 upvotes, #4 of 2026-04-06
5. Agentic-MME: What Agentic Capability Really Brings to Multimodal Intelligence? 36 upvotes, #5 of 2026-04-06
6. Token Warping Helps MLLMs Look from Nearby Viewpoints 33 upvotes, #6 of 2026-04-06
7. Test-Time Scaling Makes Overtraining Compute-Optimal 28 upvotes, #7 of 2026-04-06
8. Communicating about Space: Language-Mediated Spatial Integration Across Partial Views 20 upvotes, #8 of 2026-04-06
9. AgentHazard: A Benchmark for Evaluating Harmful Behavior in Computer-Use Agents 19 upvotes, #9 of 2026-04-06
10. Swift-SVD: Theoretical Optimality Meets Practical Efficiency in Low-Rank LLM Compression 11 upvotes, #10 of 2026-04-06
11. Xpertbench: Expert Level Tasks with Rubrics-Based Evaluation 11 upvotes, #10 of 2026-04-06
12. AgentSocialBench: Evaluating Privacy Risks in Human-Centered Agentic Social Networks 10 upvotes, #12 of 2026-04-06
13. VLMs Need Words: Vision Language Models Ignore Visual Detail In Favor of Semantic Anchors 9 upvotes, #13 of 2026-04-06
14. Do World Action Models Generalize Better than VLAs? A Robustness Study 7 upvotes, #14 of 2026-04-06
15. CoME-VL: Scaling Complementary Multi-Encoder Vision-Language Learning 7 upvotes, #14 of 2026-04-06
16. DriveDreamer-Policy: A Geometry-Grounded World-Action Model for Unified Generation and Planning 6 upvotes, #16 of 2026-04-06
17. Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generation 6 upvotes, #16 of 2026-04-06

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.