# SciFaultyQA: Benchmarking LLMs on Faulty Science Question Detection with a GAN-Inspired Approach to Synthetic Dataset Generation

arXiv: https://arxiv.org/abs/2412.11988 · HTML: https://mlsanity.com/paper/2412.11988/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Synthetic reasoning data and bootstrapping, one of 13 directions in Synthetic data and self-improvement for LLMs.
- Builds on 5 of its 15 references that are in the map.
- First posted on arXiv 16 Dec 2024.

## Placed in

- Synthetic data and self-improvement for LLMs / [Synthetic reasoning data and bootstrapping](https://mlsanity.com/synth/synthetic-reasoning-data-and-bootstrapping/)

## Builds on (most cited first, 5 of 5)

- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering](https://mlsanity.com/paper/2209.09513/) (2022)
- [Crowdsourcing Multiple Choice Science Questions](https://mlsanity.com/paper/1707.06209/) (2017)
- [Fine-grained Text Style Transfer with Diffusion-Based Language Models](https://mlsanity.com/paper/2305.19512/) (2023)

## Built on by (most cited first, 1 of 1)

- [ForensicsData: A Digital Forensics Dataset for Large Language Models](https://mlsanity.com/paper/2509.05331/) (2025)

## Abstract (arXiv)

Consider the problem: ``If one man and one woman can produce one child in one year, how many children will be produced by one woman and three men in 0.5 years?" Current large language models (LLMs) such as GPT-4o, GPT-o1-preview, and Gemini Flash frequently answer "0.5," which does not make sense. While these models sometimes acknowledge the unrealistic nature of the question, in many cases (8 out of 10 trials), they provide the nonsensical answer of "0.5 child." Additionally, temporal variation has been observed: if an LLM answers correctly once (by recognizing the faulty nature of the question), subsequent responses are more likely to also reflect this understanding. However, this is inconsistent. These types of questions have motivated us to develop a dataset of science questions, SciFaultyQA, where the questions themselves are intentionally faulty. We observed that LLMs often proceed to answer these flawed questions without recognizing their inherent issues, producing results that are logically or scientifically invalid. By analyzing such patterns, we developed a novel method for generating synthetic datasets to evaluate and benchmark the performance of various LLMs in identifying these flawed questions. We have also developed novel approaches to reduce the errors.