Daily Papers of 2026-02-16

  1. Less is Enough: Synthesizing Diverse Data in Feature Space of LLMs 225 upvotes, #1 of 2026-02-16
  2. SQuTR: A Robustness Benchmark for Spoken Query to Text Retrieval under Acoustic Noise 142 upvotes, #2 of 2026-02-16
  3. MedXIAOHE: A Comprehensive Recipe for Building Medical MLLMs 59 upvotes, #3 of 2026-02-16
  4. Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception 58 upvotes, #4 of 2026-02-16
  5. OneVision-Encoder: Codec-Aligned Sparsity as a Foundational Principle for Multimodal Intelligence 46 upvotes, #5 of 2026-02-16
  6. CoPE-VideoLM: Codec Primitives For Efficient Video Language Models 29 upvotes, #6 of 2026-02-16
  7. SemanticMoments: Training-Free Motion Similarity via Third Moment Features 21 upvotes, #7 of 2026-02-16
  8. GeoAgent: Learning to Geolocate Everywhere with Reinforced Geographic Characteristics 20 upvotes, #8 of 2026-02-16
  9. What does RL improve for Visual Reasoning? A Frankenstein-Style Analysis 14 upvotes, #9 of 2026-02-16
  10. Intelligent AI Delegation 13 upvotes, #10 of 2026-02-16
  11. ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learning 11 upvotes, #11 of 2026-02-16
  12. RLinf-Co: Reinforcement Learning-Based Sim-Real Co-Training for VLA Models 11 upvotes, #11 of 2026-02-16
  13. BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models 9 upvotes, #13 of 2026-02-16
  14. Towards Universal Video MLLMs with Attribute-Structured and Quality-Verified Instructions 8 upvotes, #14 of 2026-02-16
  15. Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution 6 upvotes, #15 of 2026-02-16
  16. DICE: Diffusion Large Language Models Excel at Generating CUDA Kernels 5 upvotes, #16 of 2026-02-16
  17. SciAgentGym: Benchmarking Multi-Step Scientific Tool-use in LLM Agents 5 upvotes, #16 of 2026-02-16
  18. Principled Synthetic Data Enables the First Scaling Laws for LLMs in Recommendation 4 upvotes, #18 of 2026-02-16
  19. Code2Worlds: Empowering Coding LLMs for 4D World Generation 4 upvotes, #18 of 2026-02-16
  20. FLAC: Maximum Entropy RL via Kinetic Energy Regularized Bridge Matching 4 upvotes, #18 of 2026-02-16
  21. Best of Both Worlds: Multimodal Reasoning and Generation via Unified Discrete Flow Matching 3 upvotes, #21 of 2026-02-16
  22. On Robustness and Chain-of-Thought Consistency of RL-Finetuned VLMs 3 upvotes, #21 of 2026-02-16
  23. Light4D: Training-Free Extreme Viewpoint 4D Video Relighting 2 upvotes, #23 of 2026-02-16
  24. TADA! Tuning Audio Diffusion Models through Activation Steering 2 upvotes, #23 of 2026-02-16
  25. Favia: Forensic Agent for Vulnerability-fix Identification and Analysis 2 upvotes, #23 of 2026-02-16
  26. Self-EvolveRec: Self-Evolving Recommender Systems with LLM-based Directional Feedback 2 upvotes, #23 of 2026-02-16
  27. Quantized Evolution Strategies: High-precision Fine-tuning of Quantized LLMs at Low-precision Cost 1 upvotes, #27 of 2026-02-16
  28. GeneralVLA: Generalizable Vision-Language-Action Models with Knowledge-Guided Trajectory Planning 1 upvotes, #27 of 2026-02-16
  29. Steer2Edit: From Activation Steering to Component-Level Editing 1 upvotes, #27 of 2026-02-16
  30. scPilot: Large Language Model Reasoning Toward Automated Single-Cell Analysis and Discovery 1 upvotes, #27 of 2026-02-16
  31. Learning Image-based Tree Crown Segmentation from Enhanced Lidar-based Pseudo-labels 1 upvotes, #27 of 2026-02-16
  32. OpenLID-v3: Improving the Precision of Closely Related Language Identification -- An Experience Report 0 upvotes, #32 of 2026-02-16

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.