Researchy Questions: A Dataset of Multi-Perspective, Decompositional Questions for LLM Web Agents
This addresses the need for more realistic and difficult benchmarks in NLP to better evaluate LLM web agents, though it is incremental as it builds on existing dataset creation methods.
The paper tackles the problem of existing QA datasets being too easy for powerful LLMs by introducing Researchy Questions, a dataset of non-factoid, multi-perspective questions derived from search engine logs, and shows that these questions are challenging for GPT-4 and benefit from decomposition techniques.
Existing question answering (QA) datasets are no longer challenging to most powerful Large Language Models (LLMs). Traditional QA benchmarks like TriviaQA, NaturalQuestions, ELI5 and HotpotQA mainly study ``known unknowns'' with clear indications of both what information is missing, and how to find it to answer the question. Hence, good performance on these benchmarks provides a false sense of security. A yet unmet need of the NLP community is a bank of non-factoid, multi-perspective questions involving a great deal of unclear information needs, i.e. ``unknown uknowns''. We claim we can find such questions in search engine logs, which is surprising because most question-intent queries are indeed factoid. We present Researchy Questions, a dataset of search engine queries tediously filtered to be non-factoid, ``decompositional'' and multi-perspective. We show that users spend a lot of ``effort'' on these questions in terms of signals like clicks and session length, and that they are also challenging for GPT-4. We also show that ``slow thinking'' answering techniques, like decomposition into sub-questions shows benefit over answering directly. We release $\sim$ 100k Researchy Questions, along with the Clueweb22 URLs that were clicked.