SurveyBench: Can LLM(-Agents) Write Academic Surveys that Align with Reader Needs?
This work addresses the need for a rigorous benchmark to assess LLM-generated academic surveys, which is incremental as it builds on existing evaluation methods by focusing on reader-aligned metrics.
The authors tackled the problem of evaluating automatically generated academic surveys by proposing SurveyBench, a fine-grained, quiz-driven evaluation framework that assesses outline quality, content quality, and non-textual richness, and found that existing LLM4Survey approaches scored on average 21% lower than human performance in content-based evaluation.
Academic survey writing, which distills vast literature into a coherent and insightful narrative, remains a labor-intensive and intellectually demanding task. While recent approaches, such as general DeepResearch agents and survey-specialized methods, can generate surveys automatically (a.k.a. LLM4Survey), their outputs often fall short of human standards and there lacks a rigorous, reader-aligned benchmark for thoroughly revealing their deficiencies. To fill the gap, we propose a fine-grained, quiz-driven evaluation framework SurveyBench, featuring (1) typical survey topics source from recent 11,343 arXiv papers and corresponding 4,947 high-quality surveys; (2) a multifaceted metric hierarchy that assesses the outline quality (e.g., coverage breadth, logical coherence), content quality (e.g., synthesis granularity, clarity of insights), and non-textual richness; and (3) a dual-mode evaluation protocol that includes content-based and quiz-based answerability tests, explicitly aligned with readers' informational needs. Results show SurveyBench effectively challenges existing LLM4Survey approaches (e.g., on average 21% lower than human in content-based evaluation).