Daily Papers of 2026-11-03

1. Video Generation Meets Agent Harness: Agentic Video Generation, Agentic World Modeling and Recursive Self-Improvement 17 upvotes, #1 of 2026-11-03
2. From Fragmentation to Fusion: A Survey and Roadmap for Omnimodal Embeddings 11 upvotes, #2 of 2026-11-03
3. Diagnosing the Reference Gap: What Visual Scaffolds Can and Cannot Fix in Multimodal Reasoning 9 upvotes, #3 of 2026-11-03
4. DecisionRank: On-Policy Distillation for Decision-Model Rerankers 4 upvotes, #4 of 2026-11-03
5. DeFT-RAG: Decomposing Multi-Hop Questions into Self-Contained Sub-Questions with Reusable Answers 3 upvotes, #5 of 2026-11-03
6. MAServeBench: Benchmarking Multi-Agent LLM Serving Systems with Hierarchical Collaboration Workloads 2 upvotes, #6 of 2026-11-03
7. Self-Improvement with Auxiliary Rewards: Learning from Verified Traces and Evaluating Transfer 2 upvotes, #6 of 2026-11-03
8. ProVe: Proposer-Verifier Decoding 1 upvotes, #8 of 2026-11-03
9. Lost in Recovery: Demystifying Recovery Behavior of Computer-Use Agents via Goal-Oriented Action Replay 1 upvotes, #8 of 2026-11-03
10. On-Policy Self-Distillation with Teacher Evolution 1 upvotes, #8 of 2026-11-03
11. TinyAstral: Transferring Self-Evolving Harnesses to Tiny Generalist Manipulation Agents 1 upvotes, #8 of 2026-11-03

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.