AI models and papers trending on October 15, 2026
The edition of this day as it was recorded: models with the likes and downloads they had then, and the most upvoted papers of the daily list.
On Thursday, October 15, 2026 the model at the top of the ranking was google/embeddinggemma-2 (feature extraction), with 135,175 downloads in 30 days and 2,290 likes. The most upvoted paper was "Player-1: Learning to Play from Synthetic Experience" from ARC Lab, Tencent, with 113 votes. 11 of the 15 leading papers had a code repository attached.
Recorded 2026-10-15 07:03 UTC · numbers are those of that moment
Papers
- 113
Player-1: Learning to Play from Synthetic Experience
ARC Lab, TencentarXiv 2610.20584 (opens in a new tab)Code (opens in a new tab)
- 106
Any Time, Any Space: A World-State-Centric Framework for Agentic Video Generation
Zhejiang UniversityarXiv 2610.20571 (opens in a new tab)Code (opens in a new tab)
- 47
LatentLoop: Scaling Looped MoE Models with Asymmetric Compute Allocation
arXiv 2610.19975 (opens in a new tab)Code (opens in a new tab)
- 44
Cortex: Consolidating Context into Parametric Memory
ByteDance SeedarXiv 2610.20399 (opens in a new tab)
- 43
ActionOPD: Distilling Decisions, Not Tokens
Tsinghua-NICS-EFCarXiv 2610.19820 (opens in a new tab)Code (opens in a new tab)
- 42
Recovering Reliable Supervision via Critique-Guided On-Policy Distillation
arXiv 2610.20572 (opens in a new tab)Code (opens in a new tab)
- 30
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory
Fudan UniversityarXiv 2610.19366 (opens in a new tab)Code (opens in a new tab)
- 29
Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents
arXiv 2610.20344 (opens in a new tab)Code (opens in a new tab)
- 29
LensErase: A Foundation Model for Erase Anything
MicrosoftarXiv 2610.18072 (opens in a new tab)Code (opens in a new tab)
- 22
KV4-RL: Unlocking Faster Asynchronous Rollouts for RLVR with INT4 KV Cache Quantization
- 22
Beyond Domain-Routed Supervision: Cross-Teacher Guarding for Multi-Teacher On-Policy Distillation
- 17
Towards Agentic Robots: Model, Data, Environment, and Harness
arXiv 2610.20442 (opens in a new tab)Code (opens in a new tab)
- 15
Hybrid Reasoning Agents
Stanford UniversityarXiv 2610.18861 (opens in a new tab)Code (opens in a new tab)
- 13
APT: Feedback-Driven Agentic Post-Training for Text-to-Image Generation
Microsoft ResearcharXiv 2610.18097 (opens in a new tab)Code (opens in a new tab)
- 12
Super-Resolution in The Right Latent Space: A Frozen Vision-Foundation Substrate
Models
- 1
Feature extraction · 135K downloads / 30 d · 2,290 likes
- 2
Text generation · 65.6K downloads / 30 d · 1,654 likes
- 3
Qwen/Qwen-Image-2.1-Turbo (opens in a new tab)
Text to image · 12.7K downloads / 30 d · 789 likes
- 4
openbmb/MiniCPM-V-4.7-35B-A3B (opens in a new tab)
Image text to text · 4,157 downloads / 30 d · 682 likes
- 5
abenzerps/Qwen-Image-2.1-Uncensored-Quantized
Text to image · 2.32M downloads / 30 d · 4,323 likes
- 6
autotrust/JEV-35B (opens in a new tab)
Text classification · 28K downloads / 30 d · 789 likes
- 7
Image to video · 1.82M downloads / 30 d · 7,830 likes
- 8
ConwayResearch/Underdog-Saluki-27B-1.0 (opens in a new tab)
Text generation · 312K downloads / 30 d · 462 likes
- 9
AtomicChat/Qwen-Image-2.1-Turbo-Uncensored-GGUF (opens in a new tab)
Text to image · 258K downloads / 30 d · 457 likes
- 10
Text to speech · 839 downloads / 30 d · 346 likes
- 11
Image text to text · 7.22M downloads / 30 d · 17.7K likes
- 12
LiquidAI/LFM2.5-24B-A2B (opens in a new tab)
Text generation · 4,067 downloads / 30 d · 329 likes
- 13
Image text to text · 114K downloads / 30 d · 339 likes
- 14
baidu/Unlimited-OCR (opens in a new tab)
Image text to text · 1.15M downloads / 30 d · 4,642 likes
- 15
Cloudflare/clef (opens in a new tab)
Image text to text · 18.3K downloads / 30 d · 2,162 likes
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.