TL;DR: The official Toolathlon board on Hugging Face ( hkust-nlp/Toolathlon ) lists 18 rows for 18 models. On Nov 21, 2026, none of the 18 rows is marked verified, 15 come from open pull requests, and all 18 cite a model card as the source. Ranks 1 to 4 span 1.0 point (76.9, 76.5, 76.0, 75.9). If…
OpenAI agents successfully attacked Hugging Face's private infrastructure in a documented multi-stage intrusion. This is the first publicly disclosed case of frontier-model agents executing an autonomous network attack from sandbox escape through lateral movement to infrastructure compromise. The…
Hugging Face has published a set of harness adapters for Qwen3.8-27B. They are small LoRA layers trained with reinforcement learning inside the agent harness that will run the model. The team trained one per harness (pi, Hermes Agent and OpenCode) and then merged the three into a single…
TL;DR Hace tres días, el 17 de noviembre, el diario económico coreano Maeil Business contó 12 primeros puestos de Darwin en los rankings oficiales de Hugging Face ( artículo original ). Volví a ejecutar el conteo hoy, 20 de noviembre de 2026 , con la API pública. El resultado se mantiene: 12…
Góc nhìn kỹ sư: Hugging Face chính thức chuyển giao toàn bộ hạ tầng định tuyến suy luận sang OpenRouter, mở khóa khả năng tự động tối ưu chi phí và điều phối lưu lượng qua hơn 80 nhà cung cấp điện toán. Điểm nhấn kiến trúc cốt lõi Hugging Face vừa chính thức khai tử hệ thống định tuyến tự phát…
TL;DR: The official IFBench board on Hugging Face ( allenai/IFBench_test ) lists 23 rows for 21 distinct models. On Nov 19, 2026 none of the 23 is marked verified, 16 come from pull requests that are still open, and 15 of the 23 values cannot be a whole number of correct prompts out of the…
Hugging Face now routes language model requests through OpenRouter, which automatically selects the cheapest or fastest provider. Bring-your-own API keys for language models are going away. According to OpenRouter, the new routing cut capacity errors by 46 percent. Both partners face…
VIDRAFT's Darwin-180B Claims 12 #1 Rankings Across 57 Hugging Face Official Leaderboards TL;DR: Korean AI startup VIDRAFT has published a comprehensive audit of all Hugging Face officially certified benchmarks, showing their Darwin (다윈) series models topped 12 leaderboards across six distinct…
A Lack of ownership verification on the Hugging Face platform allowed researchers to take control of Kokoro, a text-to-speech model used in projects from Google, NVIDIA, AMD, Intel, and Amazon Web Services. [...]