Daily Papers of 2026-04-03

  1. DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models 345 upvotes, #1 of 2026-04-03
  2. The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlook 136 upvotes, #2 of 2026-04-03
  3. Generative World Renderer 101 upvotes, #3 of 2026-04-03
  4. SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization 92 upvotes, #4 of 2026-04-03
  5. CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery 52 upvotes, #5 of 2026-04-03
  6. VOID: Video Object and Interaction Deletion 51 upvotes, #6 of 2026-04-03
  7. Steerable Visual Representations 51 upvotes, #6 of 2026-04-03
  8. EgoSim: Egocentric World Simulator for Embodied Interaction Generation 36 upvotes, #8 of 2026-04-03
  9. Therefore I am. I Think 30 upvotes, #9 of 2026-04-03
  10. LatentUM: Unleashing the Potential of Interleaved Cross-Modal Reasoning via a Latent-Space Unified Model 30 upvotes, #9 of 2026-04-03
  11. Omni-SimpleMem: Autoresearch-Guided Discovery of Lifelong Multimodal Agent Memory 29 upvotes, #11 of 2026-04-03
  12. NearID: Identity Representation Learning via Near-identity Distractors 29 upvotes, #11 of 2026-04-03
  13. UniDriveVLA: Unifying Understanding, Perception, and Action Planning for Autonomous Driving 25 upvotes, #13 of 2026-04-03
  14. ASI-Evolve: AI Accelerates AI 24 upvotes, #14 of 2026-04-03
  15. Gated Condition Injection without Multimodal Attention: Towards Controllable Linear-Attention Transformers 18 upvotes, #15 of 2026-04-03
  16. Investigating Autonomous Agent Contributions in the Wild: Activity Patterns and Code Change over Time 17 upvotes, #16 of 2026-04-03
  17. GPA: Learning GUI Process Automation from Demonstrations 16 upvotes, #17 of 2026-04-03
  18. Video Models Reason Early: Exploiting Plan Commitment for Maze Solving 14 upvotes, #18 of 2026-04-03
  19. AIBench: Evaluating Visual-Logical Consistency in Academic Illustration Generation 13 upvotes, #19 of 2026-04-03
  20. Omni123: Exploring 3D Native Foundation Models with Limited 3D Data by Unifying Text to 2D and 3D Generation 13 upvotes, #19 of 2026-04-03
  21. VideoZeroBench: Probing the Limits of Video MLLMs with Spatio-Temporal Evidence Verification 12 upvotes, #21 of 2026-04-03
  22. Tex3D: Objects as Attack Surfaces via Adversarial 3D Textures for Vision-Language-Action Models 12 upvotes, #21 of 2026-04-03
  23. Woosh: A Sound Effects Foundation Model 12 upvotes, #21 of 2026-04-03
  24. Apriel-Reasoner: RL Post-Training for General-Purpose and Efficient Reasoning 12 upvotes, #21 of 2026-04-03
  25. AutoMIA: Improved Baselines for Membership Inference Attack via Agentic Self-Exploration 11 upvotes, #25 of 2026-04-03
  26. T5Gemma-TTS Technical Report 11 upvotes, #25 of 2026-04-03
  27. MDPBench: A Benchmark for Multilingual Document Parsing in Real-World Scenarios 10 upvotes, #27 of 2026-04-03
  28. DynaVid: Learning to Generate Highly Dynamic Videos using Synthetic Motion Data 10 upvotes, #27 of 2026-04-03
  29. Forecasting Supply Chain Disruptions with Foresight Learning 9 upvotes, #29 of 2026-04-03
  30. Efficient and Principled Scientific Discovery through Bayesian Optimization: A Tutorial 9 upvotes, #29 of 2026-04-03
  31. Ask or Assume? Uncertainty-Aware Clarification-Seeking in Coding Agents 8 upvotes, #31 of 2026-04-03
  32. LinguDistill: Recovering Linguistic Ability in Vision- Language Models via Selective Cross-Modal Distillation 8 upvotes, #31 of 2026-04-03
  33. Working Notes on Late Interaction Dynamics: Analyzing Targeted Behaviors of Late Interaction Models 7 upvotes, #33 of 2026-04-03
  34. Signals: Trajectory Sampling and Triage for Agentic Interactions 7 upvotes, #33 of 2026-04-03
  35. UniRecGen: Unifying Multi-View 3D Reconstruction and Generation 7 upvotes, #33 of 2026-04-03
  36. Memory-Augmented Vision-Language Agents for Persistent and Semantically Consistent Object Captioning 6 upvotes, #36 of 2026-04-03
  37. LOME: Learning Human-Object Manipulation with Action-Conditioned Egocentric World Model 6 upvotes, #36 of 2026-04-03
  38. Executing as You Generate: Hiding Execution Latency in LLM Code Generation 6 upvotes, #36 of 2026-04-03
  39. FlowSlider: Training-Free Continuous Image Editing via Fidelity-Steering Decomposition 6 upvotes, #36 of 2026-04-03
  40. ActionParty: Multi-Subject Action Binding in Generative Video Games 6 upvotes, #36 of 2026-04-03
  41. MultiGen: Level-Design for Editable Multiplayer Worlds in Diffusion Game Engines 5 upvotes, #41 of 2026-04-03
  42. An Empirical Recipe for Universal Phone Recognition 5 upvotes, #41 of 2026-04-03
  43. Brainstacks: Cross-Domain Cognitive Capabilities via Frozen MoE-LoRA Stacks for Continual LLM Learning 5 upvotes, #41 of 2026-04-03
  44. Friends and Grandmothers in Silico: Localizing Entity Cells in Language Models 5 upvotes, #41 of 2026-04-03
  45. Automatic Image-Level Morphological Trait Annotation for Organismal Images 5 upvotes, #41 of 2026-04-03

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.