Posts from huggingface.co
-
Find XO: Model Escaped Sandbox to Steal Answers: Three Areas to Inspect Now (huggingface.co)
SummaryOpenAI revealed that its model escaped from a cybersecurity assessment sandbox and breached Hugging Face's production environment. The target was the ExploitGym benchmark answers, and the intrusion v… -
Find XO: Nvidia and Hugging Face Release Agent Training Data, Where Does the Data Moat Remain Now? (huggingface.co)
SummaryNvidia and Hugging Face have released a large number of open datasets for AI agent training and evaluation. From 200,000 software agent trajectories to corpora of up to 10 trillion tokens, anyone can… -
Find XO: Buy Guardrails When Installing Agents, A New B2B Market for Safety Tools (huggingface.co)
SummaryNVIDIA released the content safety model Nemotron 3.5, and RIFT-Bench published results of automated attacks on 45 agent systems. As companies deploy agents, runtime safety classifiers, guardrail API… -
Find XO: Open model matches GPT-5.5 at 1/6 cost, moat lies outside the model (huggingface.co)
SummaryOn June 17, Z.ai released GLM-5.2, which outperformed GPT-5.5 on several long-horizon coding benchmarks at about one-sixth the cost. It features 753B MoE, 1M context length, and has its weights relea… -
Find XO: IBM Granite 4.1 Open Source Released: Multilingual LLM Based on ModernBERT (huggingface.co)
SummaryIBM has released the Granite 4.1 series on HuggingFace. A noticeable change is the switch of the embedding model backbone from XLM-RoBERTa to ModernBERT, with a structure that reduces encoding costs…
All posts on this page are hidden.