# Perplexity Trap: PLM-Based Retrievers Overrate Low Perplexity Documents

arXiv: https://arxiv.org/abs/2503.08684 · HTML: https://mlsanity.com/paper/2503.08684/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 45 references that are in the map.
- First posted on arXiv 11 Mar 2025.

## Builds on (most cited first, 30 of 31)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Mistral 7B](https://mlsanity.com/paper/2310.06825/) (2023)
- [A Survey of Large Language Models](https://mlsanity.com/paper/2303.18223/) (2023)
- [SimCSE: Simple Contrastive Learning of Sentence Embeddings](https://mlsanity.com/paper/2104.08821/) (2021)
- [Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generation](https://mlsanity.com/paper/2305.01210/) (2023)
- [Unsupervised Dense Information Retrieval with Contrastive Learning](https://mlsanity.com/paper/2112.09118/) (2021)
- [Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval](https://mlsanity.com/paper/2007.00808/) (2020)
- [DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature](https://mlsanity.com/paper/2301.11305/) (2023)
- [MTEB: Massive Text Embedding Benchmark](https://mlsanity.com/paper/2210.07316/) (2022)
- [SPECTER: Document-level Representation Learning using Citation-informed Transformers](https://mlsanity.com/paper/2004.07180/) (2020)
- [The Curse of Recursion: Training on Generated Data Makes Models Forget](https://mlsanity.com/paper/2305.17493/) (2023)
- [Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature](https://mlsanity.com/paper/2310.05130/) (2023)
- [Overview of the TREC 2019 deep learning track](https://mlsanity.com/paper/2003.07820/) (2020)
- [Efficiently Teaching an Effective Dense Retriever with Balanced Topic Aware Sampling](https://mlsanity.com/paper/2104.06967/) (2021)
- [A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPT](https://mlsanity.com/paper/2303.04226/) (2023)
- [Unsupervised Corpus Aware Language Model Pre-training for Dense Passage Retrieval](https://mlsanity.com/paper/2108.05540/) (2021)
- [Large Language Models are Inconsistent and Biased Evaluators](https://mlsanity.com/paper/2405.01724/) (2024)
- [Dense Text Retrieval based on Pretrained Language Models: A Survey](https://mlsanity.com/paper/2211.14876/) (2022)
- [Bias and Unfairness in Information Retrieval Systems: New Challenges in the LLM Era](https://mlsanity.com/paper/2404.11457/) (2024)
- [Debiasing Graph Neural Networks via Learning Disentangled Causal Substructure](https://mlsanity.com/paper/2209.14107/) (2022)
- [Semantic Models for the First-stage Retrieval: A Comprehensive Review](https://mlsanity.com/paper/2103.04831/) (2021)
- [Neural Retrievers are Biased Towards LLM-Generated Content](https://mlsanity.com/paper/2310.20501/) (2023)
- [Learning Surrogate Losses](https://mlsanity.com/paper/1905.10108/) (2019)
- [Cocktail: A Comprehensive Information Retrieval Benchmark with LLM-Generated Documents Integration](https://mlsanity.com/paper/2405.16546/) (2024)
- [Likelihood-based Mitigation of Evaluation Bias in Large Language Models](https://mlsanity.com/paper/2402.15987/) (2024)

## Built on by (most cited first, 5 of 5)

- [LLM-Generated Fake News Induces Truth Decay in News Ecosystem: A Case Study on Neural News Recommendation](https://mlsanity.com/paper/2504.20013/) (2025)
- [LLM-Specific Utility for Retrieval-Augmented Generation](https://mlsanity.com/paper/2510.11358/) (2025)
- [How Do LLM-Generated Texts Impact Term-Based Retrieval Models?](https://mlsanity.com/paper/2508.17715/) (2025)
- [Training-Induced Bias Toward LLM-Generated Content in Dense Retrieval](https://mlsanity.com/paper/2602.10833/) (2026)
- [Do Neural Retrievers Prefer Certain Documents? Evidence of Learned Relevance Priors](https://mlsanity.com/paper/2606.02814/) (2026)

## Abstract (arXiv)

Previous studies have found that PLM-based retrieval models exhibit a preference for LLM-generated content, assigning higher relevance scores to these documents even when their semantic quality is comparable to human-written ones. This phenomenon, known as source bias, threatens the sustainable development of the information access ecosystem. However, the underlying causes of source bias remain unexplored. In this paper, we explain the process of information retrieval with a causal graph and discover that PLM-based retrievers learn perplexity features for relevance estimation, causing source bias by ranking the documents with low perplexity higher. Theoretical analysis further reveals that the phenomenon stems from the positive correlation between the gradients of the loss functions in language modeling task and retrieval task. Based on the analysis, a causal-inspired inference-time debiasing method is proposed, called Causal Diagnosis and Correction (CDC). CDC first diagnoses the bias effect of the perplexity and then separates the bias effect from the overall estimated relevance score. Experimental results across three domains demonstrate the superior debiasing effectiveness of CDC, emphasizing the validity of our proposed explanatory framework. Source codes are available at https://github.com/WhyDwelledOnAi/Perplexity-Trap.