Skip to content

Trending AI Today

AI models and papers of week 42, 2026

Archive · October 12, 2026 to October 18, 2026

7 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 7 days of model records.

The paper with the most votes this week was “Waypoint-OPD: Transferring RL-Induced Policy Shifts via Gated On-Policy Distillation”, with 666 upvotes on the list of October 16, 2026. 69 of the 80 papers recorded came with a code repository.

Papers of the week

By upvotes on their day

  1. 666

    Waypoint-OPD: Transferring RL-Induced Policy Shifts via Gated On-Policy Distillation

    Fri Oct 16arXiv 2610.21601 (opens in a new tab)Code (opens in a new tab)

  2. 577

    OmniCompass: A Holistic Benchmark for Omni-Modal Video Grounding and Understanding

    Fri Oct 16 · VCLabarXiv 2610.18829 (opens in a new tab)Code (opens in a new tab)

  3. 352

    MJev-Spatial: System One Spatial Intelligence via Native Attention Matching

    Fri Oct 16 · Mind LabarXiv 2610.22276 (opens in a new tab)Code (opens in a new tab)

  4. 315

    OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use

    Fri Oct 16 · Tongyi-MAIarXiv 2610.22200 (opens in a new tab)Code (opens in a new tab)

  5. 284

    FDE-Gym: The Last Mile of AI Delivery

    Fri Oct 16 · DeepLethearXiv 2610.22350 (opens in a new tab)Code (opens in a new tab)

  6. 201

    RoTA-Bench: Evaluating 3D Spatial Reasoning through Interactive Projection Alignment

    Fri Oct 16 · MAC-AutoMLarXiv 2610.17665 (opens in a new tab)Code (opens in a new tab)

  7. 198

    Self-Evolving Agent Harness

    Fri Oct 16 · ZGCAGIarXiv 2610.18813 (opens in a new tab)Code (opens in a new tab)

  8. 153

    MAI-Gym: An Open All-in-One Platform for GUI Agent Evaluation and RL Training across Mobile and Computer Environments

    Fri Oct 16 · Tongyi-MAIarXiv 2610.20939 (opens in a new tab)Code (opens in a new tab)

  9. 134

    Learning Where to Split: Dynamic Band Splitting for Audio Source Separation

    Fri Oct 16 · Tsinghua Machine Learning GrouparXiv 2610.14993 (opens in a new tab)Code (opens in a new tab)

  10. 118

    InfiniSplat v2: Analyzing and Improving Feed-Forward 3DGS from Any Views

    Fri Oct 16 · zju3dvarXiv 2610.20341 (opens in a new tab)Code (opens in a new tab)

  11. 113

    Player-1: Learning to Play from Synthetic Experience

    Thu Oct 15 · ARC Lab, TencentarXiv 2610.20584 (opens in a new tab)Code (opens in a new tab)

  12. 106

    Any Time, Any Space: A World-State-Centric Framework for Agentic Video Generation

    Thu Oct 15 · Zhejiang UniversityarXiv 2610.20571 (opens in a new tab)Code (opens in a new tab)

  13. 106

    Behavioral Compositionality of Sparse Autoencoder Representations

    Fri Oct 16 · OpenSAEarXiv 2610.21261 (opens in a new tab)Code (opens in a new tab)

  14. 105

    3DCodeVerse: Exploring 3D Worlds as Code

    Fri Oct 16 · University of Southern CaliforniaarXiv 2610.22341 (opens in a new tab)Code (opens in a new tab)

  15. 79

    SCOPE: Semantic Bounded Credit Optimization for Prompt Engineering in Text-to-image Generation

    Fri Oct 16 · University of RochesterarXiv 2610.22264 (opens in a new tab)Code (opens in a new tab)

  16. 77

    Speculative Forcing for Efficient Autoregressive Video Generation

    Fri Oct 16 · NAVER AI LabarXiv 2610.22347 (opens in a new tab)Code (opens in a new tab)

  17. 65

    Scaling Agentic Data for Long-Horizon Terminal Intelligence

    Tue Oct 13 · alibabaarXiv 2610.13166 (opens in a new tab)

  18. 56

    Each Reward to Its Relevant Tokens: Reward-Specific Token-Level Credit Assignment for Multi-Reward RL

    Fri Oct 16 · Mohamed Bin Zayed University of Artificial IntelligencearXiv 2610.21439 (opens in a new tab)Code (opens in a new tab)

  19. 54

    VLA-Δ: Learning Physical State Transition for Vision-Language-Action Models

    Fri Oct 16 · ZJU-OmniAIarXiv 2610.21306 (opens in a new tab)Code (opens in a new tab)

  20. 53

    GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives

    Tue Oct 13 · XPENG AIarXiv 2609.39601 (opens in a new tab)Code (opens in a new tab)

Models of the week

By days on the list, then best position

  1. google/embeddinggemma-2

    7 of 7 days · best No. 1 · Feature extraction · 158K downloads / 30 d

  2. jialinyyzz/humanizer

    7 of 7 days · best No. 2 · Text generation · 95.9K downloads / 30 d

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

    7 of 7 days · best No. 3 · Text to image · 20.7K downloads / 30 d

  4. Lightricks/LTX-2.5

    7 of 7 days · best No. 6 · Image to video · 1.99M downloads / 30 d

  5. Qwen/Qwen3.8-27B

    7 of 7 days · best No. 8 · Image text to text · 7.8M downloads / 30 d

  6. abenzerps/Qwen-Image-2.1-Uncensored-Quantized

    6 of 7 days · best No. 4 · Text to image · 2.57M downloads / 30 d

  7. AtomicChat/Qwen-Image-2.1-Turbo-Uncensored-GGUF (opens in a new tab)

    6 of 7 days · best No. 8 · Text to image · 328K downloads / 30 d

  8. openbmb/MiniCPM-V-4.7-35B-A3B (opens in a new tab)

    5 of 7 days · best No. 3 · Image text to text · 36.3K downloads / 30 d

  9. ConwayResearch/Underdog-Saluki-27B-1.0 (opens in a new tab)

    5 of 7 days · best No. 7 · Text generation · 344K downloads / 30 d

  10. baidu/Unlimited-OCR (opens in a new tab)

    5 of 7 days · best No. 10 · Image text to text · 1.48M downloads / 30 d

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

    4 of 7 days · best No. 3 · Image text to text · 18.3K downloads / 30 d

  12. fishaudio/drama-3

    4 of 7 days · best No. 4 · Text to speech · 8,193 downloads / 30 d

  13. LiquidAI/LFM2.5-24B-A2B (opens in a new tab)

    4 of 7 days · best No. 9 · Text generation · 11.9K downloads / 30 d

  14. nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2 (opens in a new tab)

    4 of 7 days · best No. 12 · Image text to text · 149K downloads / 30 d

  15. meta-models/Muse-Spark-Open-816B-A42B

    3 of 7 days · best No. 7 · Image text to text · 1,600 downloads / 30 d

  16. LiquidAI/d1-3B (opens in a new tab)

    3 of 7 days · best No. 9 · Image text to text · 13.2K downloads / 30 d

  17. black-forest-labs/flux-3-image-klein

    3 of 7 days · best No. 10 · Image to image · 46.1K downloads / 30 d

  18. convaiinnovations/laya (opens in a new tab)

    3 of 7 days · best No. 15 · Text classification · 46.2K downloads / 30 d

  19. autotrust/JEV-35B (opens in a new tab)

    2 of 7 days · best No. 6 · Text classification · 28K downloads / 30 d

  20. black-forest-labs/flux-3-image-klein-comfy

    2 of 7 days · best No. 6 · Image to image · 224K 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.