Daily Papers of 2026-07-23

  1. SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD 72 upvotes, #1 of 2026-07-23
  2. Self Gradient Forcing: Native Long Video Extrapolation 34 upvotes, #2 of 2026-07-23
  3. Beyond Relevance-Centric Retrieval: Rubric-Oriented Document Set Selection and Ranking 31 upvotes, #3 of 2026-07-23
  4. An Exam for Active Observers 25 upvotes, #4 of 2026-07-23
  5. Scaling Laws for Hypernetwork-Based Knowledge Injection in Large Language Models 17 upvotes, #5 of 2026-07-23
  6. Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment 16 upvotes, #6 of 2026-07-23
  7. Beyond Euclidean Clipping: Overcoming Exploration Collapse in LLM RL via Riemannian Isometric Policy Optimization 14 upvotes, #7 of 2026-07-23
  8. DocOps: A Verifiable Benchmark for Autonomous Agents in Complex Document Operations 9 upvotes, #8 of 2026-07-23
  9. ATSplat: Compact Feed-forward 3D Gaussian Splatting with Adaptive Token Expansion 9 upvotes, #8 of 2026-07-23
  10. FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation 8 upvotes, #10 of 2026-07-23
  11. Train the Model, Not the Reader: Decodability Supervision for Verifiable Activation Explanations 7 upvotes, #11 of 2026-07-23
  12. ICAE-Bench: Evaluating Coding Agents as Interactive Project Builders 6 upvotes, #12 of 2026-07-23
  13. SLPO: Scaling Latent Reasoning via a Surrogate Policy 5 upvotes, #13 of 2026-07-23
  14. Trace: A Taxonomy-Guided Environment for Multidomain Visual Reasoning 5 upvotes, #13 of 2026-07-23
  15. Differentiable Logic Gate Networks for Low-Latency EEG Classification on Edge Devices 4 upvotes, #15 of 2026-07-23
  16. G-MAD: A Game-Based Data Generation Framework for Multi-View RGB-T Aerial Object Detection 4 upvotes, #15 of 2026-07-23
  17. SeededGrasp: Language-Guided Grasping in Complex Scenes with Multiple Embodiments 4 upvotes, #15 of 2026-07-23
  18. SLAM in Low-Light Environments: Project Report 2 upvotes, #18 of 2026-07-23
  19. ENTRAP-VL: A Taxonomic Probe for Dual Contextual Entrainment in Vision-Language Models 2 upvotes, #18 of 2026-07-23
  20. Moving Alphabet: A Controlled Study of Training Data for Text-to-Video Generation 1 upvotes, #20 of 2026-07-23

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.