ReFACT: A Benchmark for Scientific Confabulation Detection with Positional Error Annotations
Abstract
A benchmark called ReFACT evaluates the detection, localization, and correction of scientific confabulation in LLMs, revealing their limited accuracy.
Large Language Models (LLMs) frequently confabulate scientific facts, severely undermining their trustworthiness. Addressing this challenge requires benchmarks that go beyond binary factuality and enable fine-grained evaluation. We introduce ReFACT (Reddit False And Correct Texts), a benchmark of 1,001 expert-annotated question-answer pairs spanning diverse scientific domains for the detection of scientific confabulation. Each instance includes both a scientifically correct answer and a non-factual counterpart annotated with precise error spans and error types. ReFACT enables multi-stage evaluation: (1) confabulation detection, (2) fine-grained error localization, and (3) correction. We benchmark 9 state-of-the-art LLMs, revealing limited performance (about 50 percent accuracy). Even top models such as GPT-4o fail to distinguish factual from confabulated scientific answers, raising concerns about the reliability of LLM-as-judge evaluation paradigms. Our findings highlight the need for fine-grained, human-validated benchmarks to detect and correct scientific confabulation in domain-specific contexts. The dataset is available at: https://github.com/ddz5431/ReFACT
Models citing this paper 0
No model linking this paper
Datasets citing this paper 1
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper