Junxian Li

Junxian Li on Hugging Face Daily Papers: 12 papers, 2 in the top 3 of their day, 326 upvotes.

  1. Fewer Tokens, More Self-Teaching: On-Policy Self-Distillation for Extreme Visual Token Reduction 8 upvotes, #53 of 2026-09-29
  2. Just MLPs: Efficient Visual State Reconstruction for Multimodal Language Models 30 upvotes, #30 of 2026-09-29
  3. δ-mem: Efficient Online Memory for Large Language Models 119 upvotes, #3 of 2026-05-13
  4. SlowBA: An efficiency backdoor attack towards VLM-based GUI agents 1 upvotes, #39 of 2026-03-10
  5. PlanViz: Evaluating Planning-Oriented Image Generation and Editing for Computer-Use Tasks 5 upvotes, #29 of 2026-02-09
  6. AI for Service: Proactive Assistance with AI Glasses 71 upvotes, #4 of 2025-10-17
  7. Mol-R1: Towards Explicit Long-CoT Reasoning in Molecule Discovery 37 upvotes, #4 of 2025-08-14
  8. IAG: Input-aware Backdoor Attack on VLMs for Visual Grounding 7 upvotes, #14 of 2025-08-14
  9. Control-R: Towards controllable test-time scaling 3 upvotes, #35 of 2025-06-04
  10. TOMG-Bench: Evaluating LLMs on Text-based Open Molecule Generation 4 upvotes, #14 of 2024-12-20
  11. Critic-V: VLM Critics Help Catch VLM Errors in Multimodal Reasoning 28 upvotes, #1 of 2024-11-29
  12. Seeing and Understanding: Bridging Vision with Chemical Knowledge Via ChemVLM 19 upvotes, #6 of 2024-08-14

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.