Daily Papers of 2026-03-02

  1. dLLM: Simple Diffusion Language Modeling 123 upvotes, #1 of 2026-03-02
  2. CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation 80 upvotes, #2 of 2026-03-02
  3. Enhancing Spatial Understanding in Image Generation via Reward Modeling 50 upvotes, #3 of 2026-03-02
  4. Recovered in Translation: Efficient Pipeline for Automated Translation of Benchmarks and Datasets 39 upvotes, #4 of 2026-03-02
  5. Mode Seeking meets Mean Seeking for Fast Long Video Generation 38 upvotes, #5 of 2026-03-02
  6. LK Losses: Direct Acceptance Rate Optimization for Speculative Decoding 18 upvotes, #6 of 2026-03-02
  7. CiteAudit: You Cited It, But Did You Read It? A Benchmark for Verifying Scientific References in the LLM Era 16 upvotes, #7 of 2026-03-02
  8. How to Take a Memorable Picture? Empowering Users with Actionable Feedback 14 upvotes, #8 of 2026-03-02
  9. Compositional Generalization Requires Linear, Orthogonal Representations in Vision Embedding Models 14 upvotes, #8 of 2026-03-02
  10. InfoNCE Induces Gaussian Distribution 12 upvotes, #10 of 2026-03-02
  11. LongVideo-R1: Smart Navigation for Low-cost Long Video Understanding 10 upvotes, #11 of 2026-03-02
  12. Ref-Adv: Exploring MLLM Visual Reasoning in Referring Expression Tasks 10 upvotes, #11 of 2026-03-02
  13. Accelerating Masked Image Generation by Learning Latent Controlled Dynamics 8 upvotes, #13 of 2026-03-02
  14. Memory Caching: RNNs with Growing Memory 8 upvotes, #13 of 2026-03-02
  15. SenCache: Accelerating Diffusion Model Inference via Sensitivity-Aware Caching 7 upvotes, #15 of 2026-03-02
  16. Shared Nature, Unique Nurture: PRISM for Pluralistic Reasoning via In-context Structure Modeling 4 upvotes, #16 of 2026-03-02
  17. DUET-VLM: Dual stage Unified Efficient Token reduction for VLM Training and Inference 3 upvotes, #17 of 2026-03-02
  18. Vectorizing the Trie: Efficient Constrained Decoding for LLM-based Generative Retrieval on Accelerators 3 upvotes, #17 of 2026-03-02
  19. DLEBench: Evaluating Small-scale Object Editing Ability for Instruction-based Image Editing Model 3 upvotes, #17 of 2026-03-02
  20. CL4SE: A Context Learning Benchmark For Software Engineering Tasks 2 upvotes, #20 of 2026-03-02
  21. Reinforcement-aware Knowledge Distillation for LLM Reasoning 1 upvotes, #21 of 2026-03-02
  22. Cognitive Models and AI Algorithms Provide Templates for Designing Language Agents 1 upvotes, #21 of 2026-03-02

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.