一句话：提出可插拔的不完美用户模拟工具 FallibleUser，通过答案改写与严格判定，揭示 oracle 用户评估会高估交互式 AI 系统鲁棒性。
交互式 LLM 与 agent 评估常依赖 oracle 用户或知情模拟器，它们总能给出完整明确的答案，从而可能高估系统面对真实用户遗漏信息、模糊回答或不清指代时的鲁棒性。作者提出 FallibleUser：一个可插拔的不完美用户模拟工具包。它保留原有 oracle 用户流程，但在答案返回系统前插入可配置改写层，支持槽位遗漏、降低具体性和指代歧义三种算子；并配合严格的答案后 Q&A 级命中判定，只有最终交互引出完整、具体且无歧义的信息才计成功。在需求获取场景中，该工具表明 oracle 评估会显著高估同一访谈 agent 的鲁棒性。
Interactive LLM and agent evaluations often rely on oracle users or ground-truth-aware simulators that provide complete and unambiguous answers once the system asks a relevant question. This idealized assumption makes evaluation scalable and reproducible, but can overestimate robustness when real users omit information, give vague answers, or use unclear references. We present FallibleUser, a pluggable fallible-user simulation toolkit for evaluating interactive AI systems under imperfect user answers. FallibleUser preserves the original oracle-user pipeline, but inserts a configurable answer-rewriting layer before the answer is returned to the evaluated system. The current toolkit supports three representative fallibility operators: slot omission, specificity reduction, and referential ambiguity. It also pairs rewriting with a strict post-answer Q&A-level hit judge, so that a target is credited only when the final interaction actually elicits complete, specific, and unambiguous information. We demonstrate FallibleUser in a requirements-elicitation evaluation setting and show that oracle-user evaluation can substantially overestimate the robustness of the same interviewer agent. FallibleUser is open-sourced at https://doi.org/10.5281/zenodo.20136882, with a demonstration video available at https://youtu.be/T8-I9pb8rAI.
@inproceedings{paperbot12465,
title = {FallibleUser: An Imperfect User Agent for Realistic Evaluation of Interactive AI Systems},
author = {Yaotian Yang and Zhi Jin and Dongming Jin and Xiaohong Chen},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3834622}
}