两大AI评委给Agent回答高分0.85，但Agent从未打开文件，GroundEval揭穿这巨大评分偏差。
Article URL: https://tenureai.dev/writing/llm-as-judge-became-the-default-for-agent-evaluation/ Comments URL: https://news.ycombinator.com/item?id=486…
一个开源项目教你用简洁指标衡量AI代理的真实价值，告别模糊评估。
Article URL: https://github.com/Idank96/agent-panorama Comments URL: https://news.ycombinator.com/item?id=48531114 Points: 1 # Comments: 1
模拟真实企业事件的动态合成数据集生成器，支持多种部署方式，专为AI agent性能评估而设计。
Article URL: https://github.com/aeriesec/orgforge Comments URL: https://news.ycombinator.com/item?id=48494178 Points: 2 # Comments: 0
对照实验揭示GAIA基准中脚手架如何混淆模型真实能力，为agent评估提供关键修正。
arXiv:2606.08529v1 Announce Type: cross Abstract: Published agent capability scores conflate what a model can do with what its scaffold lets it do, an…
最新基准AgentProcessBench，精准诊断工具使用Agent每一步的过程质量，从结果到过程全面评估Agent可靠性
arXiv:2603.14465v2 Announce Type: replace Abstract: While Large Language Models (LLMs) have evolved into tool-using agents, they remain brittle in lon…
反事实评估直击临床大模型盲区，暴露出传统测试无法察觉的能力不均衡，为AI安全应用画红线
arXiv:2605.30590v1 Announce Type: cross Abstract: Two clinical AI systems can score nearly identically on coverage-based rubrics yet behave radically …
AI智能体在个性化智能家居环境中的表现如何？全新基准PersonalHomeBench帮你科学评估。
arXiv:2604.16813v3 Announce Type: replace Abstract: Agentic AI systems are rapidly advancing toward real-world applications, yet their readiness in co…