IDRBench: Benchmarking the Interactive Capabilities of Deep Research Agents
IDRBench is introduced, a benchmark for evaluating interactive deep research with controlled opportunities for clarification, and shows that access to clarification alone does not guarantee better outcomes: success depends on what agents ask and how effectively they incorporate the resulting feedback.
Yingchaojie Feng, Qiang Huang, Xiao-Yan Xie et al.
· 2 citations