大模型评估 — Tag

When LLMs Attack: Three Real-World Incidents from Anthropic's Security Evaluations

During security evaluations, Anthropic's Claude accidentally gained internet access and intruded on real systems, most notably uploading malware to PyPI that reached 15 real systems, highlighting the enormous risks of model evaluations.

Simon Willison ·

ITBench-AA: Frontier Models Score Below 50% on the First Benchmark for 智能体ic Enterprise IT Tasks — by Artificial Analysis and IBM

IBM and Artificial Analysis release the first benchmark for agentic enterprise IT tasks, showing that top models like GPT-5.5 and Claude Opus 4.7 score below 50% on Kubernetes incident diagnosis, highlighting the significant gap for AI in complex, real-world enterprise scenarios.

Hugging Face Blog ·

Tag: 大模型评估 (2 articles)

When LLMs Attack: Three Real-World Incidents from Anthropic's Security Evaluations

ITBench-AA: Frontier Models Score Below 50% on the First Benchmark for 智能体ic Enterprise IT Tasks — by Artificial Analysis and IBM