# Unsupervised Distractor Generation via Large Language Model Distilling and Counterfactual Contrastive Decoding

arXiv: https://arxiv.org/abs/2406.01306 · HTML: https://mlsanity.com/paper/2406.01306/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 26 references that are in the map.
- First posted on arXiv 3 Jun 2024.

## Builds on (most cited first, 18 of 18)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [A Diversity-Promoting Objective Function for Neural Conversation Models](https://mlsanity.com/paper/1510.03055/) (2015)
- [ChatGPT Outperforms Crowd-Workers for Text-Annotation Tasks](https://mlsanity.com/paper/2303.15056/) (2023)
- [RACE: Large-scale ReAding Comprehension Dataset From Examinations](https://mlsanity.com/paper/1704.04683/) (2017)
- [Want To Reduce Labeling Cost? GPT-3 Can Help](https://mlsanity.com/paper/2108.13487/) (2021)
- [AnnoLLM: Making Large Language Models to Be Better Crowdsourced Annotators](https://mlsanity.com/paper/2303.16854/) (2023)
- [DREAM: A Challenge Dataset and Models for Dialogue-Based Reading Comprehension](https://mlsanity.com/paper/1902.00164/) (2019)
- [Ask Me Anything: A simple strategy for prompting language models](https://mlsanity.com/paper/2210.02441/) (2022)
- [Language Models in the Loop: Incorporating Prompting into Weak Supervision](https://mlsanity.com/paper/2205.02318/) (2022)
- [LLM-powered Data Augmentation for Enhanced Cross-lingual Performance](https://mlsanity.com/paper/2305.14288/) (2023)
- [UDAPDR: Unsupervised Domain Adaptation via LLM Prompting and Distillation of Rerankers](https://mlsanity.com/paper/2303.00807/) (2023)
- [A Comparative Study of AI-Generated (GPT-4) and Human-crafted MCQs in Programming Education](https://mlsanity.com/paper/2312.03173/) (2023)
- [Automatic Distractor Generation for Multiple Choice Questions in Standard Tests](https://mlsanity.com/paper/2011.13100/) (2020)
- [Generating Distractors for Reading Comprehension Questions from Real Examinations](https://mlsanity.com/paper/1809.02768/) (2018)
- [Knowledge-Driven Distractor Generation for Cloze-style Multiple Choice Questions](https://mlsanity.com/paper/2004.09853/) (2020)
- [Distractor generation for multiple-choice questions with predictive prompting and large language models](https://mlsanity.com/paper/2307.16338/) (2023)
- [Co-Attention Hierarchical Network: Generating Coherent Long Distractors for Reading Comprehension](https://mlsanity.com/paper/1911.08648/) (2019)
- [Improving Machine Reading Comprehension with Single-choice Decision and Transfer Learning](https://mlsanity.com/paper/2011.03292/) (2020)

## Built on by (most cited first, 11 of 11)

- [Merge, Ensemble, and Cooperate! A Survey on Collaborative Strategies in the Era of Large Language Models](https://mlsanity.com/paper/2407.06089/) (2024)
- [HumanVBench: Probing Human-Centric Video Understanding in MLLMs with Automatically Synthesized Benchmarks](https://mlsanity.com/paper/2412.17574/) (2024)
- [Generating Plausible Distractors for Multiple-Choice Questions via Student Choice Prediction](https://mlsanity.com/paper/2501.13125/) (2025)
- [DiVERT: Distractor Generation with Variational Errors Represented as Text for Math Multiple-choice Questions](https://mlsanity.com/paper/2406.19356/) (2024)
- [Wrong Answers Can Also Be Useful: PlausibleQA -- A Large-Scale QA Dataset with Answer Plausibility Scores](https://mlsanity.com/paper/2502.16358/) (2025)
- [Mitigating Easy Option Bias in Multiple-Choice Question Answering](https://mlsanity.com/paper/2508.13428/) (2025)
- [Lost in the Passage: Passage-level In-context Learning Does Not Necessarily Need a "Passage"](https://mlsanity.com/paper/2502.10634/) (2025)
- [The Ever-Evolving Science Exam](https://mlsanity.com/paper/2507.16514/) (2025)
- [D-GEN: Automatic Distractor Generation and Evaluation for Reliable Assessment of Generative Model](https://mlsanity.com/paper/2504.13439/) (2025)
- [DualReward: A Dynamic Reinforcement Learning Framework for Cloze Tests Distractor Generation](https://mlsanity.com/paper/2507.11875/) (2025)
- [Corpus Prevalence of Multiple-Choice Question Options](https://mlsanity.com/paper/2602.17377/) (2026)

## Abstract (arXiv)

Within the context of reading comprehension, the task of Distractor Generation (DG) aims to generate several incorrect options to confuse readers. Traditional supervised methods for DG rely heavily on expensive human-annotated distractor labels. In this paper, we propose an unsupervised DG framework, leveraging Large Language Models (LLMs) as cost-effective annotators to enhance the DG capability of smaller student models. Specially, to perform knowledge distilling, we propose a dual task training strategy that integrates pseudo distractors from LLMs and the original answer in-formation as the objective targets with a two-stage training process. Moreover, we devise a counterfactual contrastive decoding mechanism for increasing the distracting capability of the DG model. Experiments show that our unsupervised generation method with Bart-base greatly surpasses GPT-3.5-turbo performance with only 200 times fewer model parameters. Our proposed unsupervised DG method offers a cost-effective framework for practical reading comprehension applications, without the need of laborious distractor annotation and costly large-size models