Skip to content

Trending AI Today

AI models and papers of week 40, 2026

Archive · September 28, 2026 to October 4, 2026

6 recorded days folded into one page: the 20 most upvoted papers of the week and the models that held a place on the trending list, from 1 day of model records.

The paper with the most votes this week was “Raven: The Harness of Harnesses for Composable Agentic Intelligence” from EverMind, with 515 upvotes on the list of September 30, 2026. 53 of the 75 papers recorded came with a code repository.

Papers of the week

By upvotes on their day

  1. 515

    Raven: The Harness of Harnesses for Composable Agentic Intelligence

    Wed Sep 30 · EverMindarXiv 2609.33439 (opens in a new tab)Code (opens in a new tab)

  2. 400

    MaLiang-Harness: A Programmable Path to Image and Video Generation

    Wed Sep 30 · National University of SingaporearXiv 2609.34309 (opens in a new tab)Code (opens in a new tab)

  3. 372

    In-Context Learning for Robots: Methods and Applications

    Wed Sep 30 · Knowin AIarXiv 2609.36012 (opens in a new tab)Code (opens in a new tab)

  4. 372

    The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation

    Thu Oct 1arXiv 2609.36484 (opens in a new tab)Code (opens in a new tab)

  5. 335

    False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents

    Thu Oct 1 · Rutgers UniversityarXiv 2609.39102 (opens in a new tab)

  6. 309

    Scaling Properties of Same-Family On-Policy Distillation

    Wed Sep 30 · Zhejiang UniversityarXiv 2609.32722 (opens in a new tab)

  7. 282

    Omni-IO Skills: Harnessing Your Agent Omni-Native

    Wed Sep 30 · National University of SingaporearXiv 2609.31847 (opens in a new tab)Code (opens in a new tab)

  8. 273

    UniEvo-VL: An On-policy Self-Distillation Training Recipe for Multimodal Model Self-improvement

    Thu Oct 1 · Stanford NLParXiv 2609.38721 (opens in a new tab)

  9. 271

    Post-Training Leaves Behavioral Shadows on Unrelated Decisions

    Tue Sep 29 · Peking UniversityarXiv 2609.29233 (opens in a new tab)Code (opens in a new tab)

  10. 241

    YuE2: Unifying Symbolic and Audio Music Generation at Frontier Quality

    Tue Sep 29 · Multimodal Art ProjectionarXiv 2609.33757 (opens in a new tab)Code (opens in a new tab)

  11. 215

    VisionHOPE: Visual Backbones as Self-Modifying Learning Systems

    Tue Sep 29 · Mininglamp TechnologyarXiv 2609.33325 (opens in a new tab)Code (opens in a new tab)

  12. 209

    Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence

    Thu Oct 1 · AmazonarXiv 2609.34563 (opens in a new tab)Code (opens in a new tab)

  13. 176

    Beyond Teacher Assignment: Domain-Normalized Multi-Teacher On-Policy Distillation

    Tue Sep 29 · Nanyang Technological UniversityarXiv 2609.35347 (opens in a new tab)Code (opens in a new tab)

  14. 159

    PanoVLN: Towards Effective Panoramic Vision-and-Language Navigation

    Wed Sep 30 · Zhejiang UniversityarXiv 2609.34759 (opens in a new tab)Code (opens in a new tab)

  15. 158

    FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders

    Mon Sep 28 · USC Physical Superintelligence (PSI) LabarXiv 2609.31620 (opens in a new tab)Code (opens in a new tab)

  16. 152

    OneStreamer: Unifying Perception, Memory, and Proactive Response in Streaming Video Interaction

    Fri Oct 2 · Nanjing UniversityarXiv 2610.01762 (opens in a new tab)Code (opens in a new tab)

  17. 143

    VoxMem: Benchmarking Multimodal Memory in Large Audio Language Models

    Wed Sep 30 · The University of MelbournearXiv 2609.32607 (opens in a new tab)Code (opens in a new tab)

  18. 130

    On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics

    Fri Oct 2 · University of CambridgearXiv 2609.35259 (opens in a new tab)

  19. 128

    Groupwise Agentic Grading and Advantage Redistribution for Code Agent RL

    Tue Sep 29 · Xiaomi MiMoarXiv 2609.32577 (opens in a new tab)

  20. 127

    LEGO-Anything: Coding Agents for 3D Scene Reconstruction

    Wed Sep 30 · Amazon Web ServicesarXiv 2609.36380 (opens in a new tab)

Models of the week

By days on the list, then best position

  1. Cloudflare/clef (opens in a new tab)

    1 of 1 day · best No. 1 · Image text to text · 824 downloads / 30 d

  2. abenzerps/Qwen-Image-2.1-Uncensored-GGUF (opens in a new tab)

    1 of 1 day · best No. 2 · Text to image · 1.38M downloads / 30 d

  3. Contrastive-LM/CLM-v0.1-8B (opens in a new tab)

    1 of 1 day · best No. 3 · Text ranking · 2,951 downloads / 30 d

  4. Lightricks/LTX-2.5

    1 of 1 day · best No. 4 · Image to video · 1.63M downloads / 30 d

  5. Qwen/Qwen-Image-2.1 (opens in a new tab)

    1 of 1 day · best No. 5 · Text to image · 81.7K downloads / 30 d

  6. SupersonicLabs/Julia-1 (opens in a new tab)

    1 of 1 day · best No. 6 · Text classification · 2,909 downloads / 30 d

  7. Qwen/Qwen3.8-27B

    1 of 1 day · best No. 7 · Image text to text · 6.9M downloads / 30 d

  8. Cloudflare/clef-flash (opens in a new tab)

    1 of 1 day · best No. 8 · Image text to text · 1,303 downloads / 30 d

  9. PSRben/VisionHOPE (opens in a new tab)

    1 of 1 day · best No. 9 · Image classification · 1,279 downloads / 30 d

  10. Viggle/Qwen-Image-2.1-viggle-turbo (opens in a new tab)

    1 of 1 day · best No. 10 · Text to image · 241K downloads / 30 d

  11. nvidia/Nemotron-3-Diarization (opens in a new tab)

    1 of 1 day · best No. 11 · Voice activity detection · 44.4K downloads / 30 d

  12. orcarouter/OrcaSAQ-2-Cyber-27B-Uncensored-GGUF (opens in a new tab)

    1 of 1 day · best No. 12 · Text generation · 9,533 downloads / 30 d

  13. prism-ml/Ternary-Bonsai-2-27B-gguf (opens in a new tab)

    1 of 1 day · best No. 13 · Text generation · 3.87M downloads / 30 d

  14. TaichuAI/ZDTaichu5.0-9B (opens in a new tab)

    1 of 1 day · best No. 14 · Image text to text · 12.4K downloads / 30 d

  15. ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF (opens in a new tab)

    1 of 1 day · best No. 15 · Image text to text · 261K downloads / 30 d

About the archive

How far back does the archive go?

Model lists exist from the first day the collector ran: Hugging Face does not publish past trending lists, so they cannot be rebuilt. Paper lists for earlier days were rebuilt from the Hugging Face daily papers archive, and those days are marked "papers only".

What is kept for each day?

The leading models with their task, likes and downloads on that day, the models held back, and the leading papers with their upvotes. Abstracts are not duplicated in the archive; each paper links to arXiv.

Do archived numbers change?

No. A day’s record is the last collection of that day and is not rewritten afterwards. Likes and downloads you see on an archived day are the values on that day, not today’s.

How are weeks defined?

ISO 8601 weeks: Monday to Sunday, numbered so that week 1 contains the first Thursday of the year. The address /week/2026-40 is week 40 of 2026.