Daily Papers of 2026-04-28

  1. From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company 99 upvotes, #1 of 2026-04-28
  2. World-R1: Reinforcing 3D Constraints for Text-to-Video Generation 34 upvotes, #2 of 2026-04-28
  3. ReVSI: Rebuilding Visual Spatial Intelligence Evaluation for Accurate Assessment of VLM 3D Reasoning 31 upvotes, #3 of 2026-04-28
  4. Vision-Language-Action Safety: Threats, Challenges, Evaluations, and Mechanisms 26 upvotes, #4 of 2026-04-28
  5. Why Fine-Tuning Encourages Hallucinations and How to Fix It 25 upvotes, #5 of 2026-04-28
  6. ClawMark: A Living-World Benchmark for Multi-Turn, Multi-Day, Multimodal Coworker Agents 20 upvotes, #6 of 2026-04-28
  7. For-Value: Efficient Forward-Only Data Valuation for finetuning LLMs and VLMs 19 upvotes, #7 of 2026-04-28
  8. Sapiens2 16 upvotes, #8 of 2026-04-28
  9. Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis 14 upvotes, #9 of 2026-04-28
  10. Towards Understanding the Robustness of Sparse Autoencoders 11 upvotes, #10 of 2026-04-28
  11. SketchVLM: Vision language models can annotate images to explain thoughts and guide users 11 upvotes, #10 of 2026-04-28
  12. UniGeo: Unifying Geometric Guidance for Camera-Controllable Image Editing via Video Models 9 upvotes, #12 of 2026-04-28
  13. Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment 9 upvotes, #12 of 2026-04-28
  14. How Much Is One Recurrence Worth? Iso-Depth Scaling Laws for Looped Language Models 9 upvotes, #12 of 2026-04-28
  15. Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation 8 upvotes, #15 of 2026-04-28
  16. Efficient Agent Evaluation via Diversity-Guided User Simulation 7 upvotes, #16 of 2026-04-28
  17. TexOCR: Advancing Document OCR Models for Compilable Page-to-LaTeX Reconstruction 6 upvotes, #17 of 2026-04-28
  18. Learning to Identify Out-of-Distribution Objects for 3D LiDAR Anomaly Segmentation 6 upvotes, #17 of 2026-04-28
  19. PageGuide: Browser extension to assist users in navigating a webpage and locating information 6 upvotes, #17 of 2026-04-28
  20. ATTN-FIQA: Interpretable Attention-based Face Image Quality Assessment with Vision Transformers 4 upvotes, #20 of 2026-04-28
  21. RaV-IDP: A Reconstruction-as-Validation Framework for Faithful Intelligent Document Processing 4 upvotes, #20 of 2026-04-28
  22. Stabilizing Efficient Reasoning with Step-Level Advantage Selection 4 upvotes, #20 of 2026-04-28
  23. IndustryAssetEQA: A Neurosymbolic Operational Intelligence System for Embodied Question Answering in Industrial Asset Maintenance 3 upvotes, #23 of 2026-04-28
  24. OmniShotCut: Holistic Relational Shot Boundary Detection with Shot-Query Transformer 3 upvotes, #23 of 2026-04-28
  25. Disentangled Robot Learning via Separate Forward and Inverse Dynamics Pretraining 2 upvotes, #25 of 2026-04-28
  26. EX-FIQA: Leveraging Intermediate Early eXit Representations from Vision Transformers for Face Image Quality Assessment 2 upvotes, #25 of 2026-04-28
  27. Discovering Agentic Safety Specifications from 1-Bit Danger Signals 2 upvotes, #25 of 2026-04-28
  28. Improving Robustness of Tabular Retrieval via Representational Stability 2 upvotes, #25 of 2026-04-28
  29. Improving Vision-language Models with Perception-centric Process Reward Models 2 upvotes, #25 of 2026-04-28
  30. ProEval: Proactive Failure Discovery and Efficient Performance Estimation for Generative AI Evaluation 1 upvotes, #30 of 2026-04-28
  31. Personality Shapes Gender Bias in Persona-Conditioned LLM Narratives Across English and Hindi: An Empirical Investigation 1 upvotes, #30 of 2026-04-28
  32. Credal Concept Bottleneck Models for Epistemic-Aleatoric Uncertainty Decomposition 1 upvotes, #30 of 2026-04-28
  33. Quantum Kernel Advantage over Classical Collapse in Medical Foundation Model Embeddings 1 upvotes, #30 of 2026-04-28
  34. Stochastic KV Routing: Enabling Adaptive Depth-Wise Cache Sharing 8 upvotes, #34 of 2026-04-28
  35. Zero-to-CAD: Agentic Synthesis of Interpretable CAD Programs at Million-Scale Without Real Data 9 upvotes, #34 of 2026-04-28

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.