评分 7 · 来源：cs.AI updates on arXiv.org · 发布于 2026-04-14
评分依据：Long-context autonomous agent benchmark addressing gap between single-capability tests and real-world multi-step scenarios. 1M-token context is forward-looking.
AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts
发布
采集
学术前沿 7.0 分
— Long-context autonomous agent benchmark addressing gap between single-capability tests and real-world multi-step scenarios. 1M-token context is forward-looking.