# PaperAsk: A Benchmark for Reliability Evaluation of LLMs in Paper Search and Reading

arXiv: https://arxiv.org/abs/2510.22242 · HTML: https://mlsanity.com/paper/2510.22242/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 36 references that are in the map.
- First posted on arXiv 25 Oct 2025.

## Builds on (most cited first, 20 of 20)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection](https://mlsanity.com/paper/2302.12173/) (2023)
- [A Dataset of Information-Seeking Questions and Answers Anchored in Research Papers](https://mlsanity.com/paper/2105.03011/) (2021)
- [LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods](https://mlsanity.com/paper/2412.05579/) (2024)
- [Hallucination is Inevitable: An Innate Limitation of Large Language Models](https://mlsanity.com/paper/2401.11817/) (2024)
- [Poisoning Web-Scale Training Datasets is Practical](https://mlsanity.com/paper/2302.10149/) (2023)
- [Benchmarking and Defending Against Indirect Prompt Injection Attacks on Large Language Models](https://mlsanity.com/paper/2312.14197/) (2023)
- [MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent](https://mlsanity.com/paper/2507.02259/) (2025)
- [Why Language Models Hallucinate](https://mlsanity.com/paper/2509.04664/) (2025)
- [Limitations of the LLM-as-a-Judge Approach for Evaluating LLM Outputs in Expert Knowledge Tasks](https://mlsanity.com/paper/2410.20266/) (2024)
- [Agentic Reasoning: A Streamlined Framework for Enhancing LLM Reasoning with Agentic Tools](https://mlsanity.com/paper/2502.04644/) (2025)
- [Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety](https://mlsanity.com/paper/2502.05206/) (2025)
- [PaSa: An LLM Agent for Comprehensive Academic Paper Search](https://mlsanity.com/paper/2501.10120/) (2025)
- [Tool Learning in the Wild: Empowering Language Models as Automatic Tool Agents](https://mlsanity.com/paper/2405.16533/) (2024)
- [LitSearch: A Retrieval Benchmark for Scientific Literature Search](https://mlsanity.com/paper/2407.18940/) (2024)
- [LLM4Rerank: LLM-based Auto-Reranking Framework for Recommendations](https://mlsanity.com/paper/2406.12433/) (2024)
- [Small Models, Big Insights: Leveraging Slim Proxy Models To Decide When and What to Retrieve for LLMs](https://mlsanity.com/paper/2402.12052/) (2024)
- [How Deep Do Large Language Models Internalize Scientific Literature and Citation Practices?](https://mlsanity.com/paper/2504.02767/) (2025)
- [ADMIT: Few-shot Knowledge Poisoning Attacks on RAG-based Fact Checking](https://mlsanity.com/paper/2510.13842/) (2025)

## Built on by (most cited first, 3 of 3)

- [Structurally Human, Semantically Biased: Detecting LLM-Generated References with Embeddings and GNNs](https://mlsanity.com/paper/2601.20704/) (2026)
- [Do AI chatbots find what experts would? Effects of model, user role, and sample size on study retrieval for medical questions](https://mlsanity.com/paper/2608.13786/) (2026)
- [VEX-Bench: Benchmarking Verification Complexity of LLM-Generated Misinformation](https://mlsanity.com/paper/2609.35028/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) increasingly serve as research assistants, yet their reliability in scholarly tasks remains under-evaluated. In this work, we introduce PaperAsk, a benchmark that systematically evaluates LLMs across four key research tasks: citation retrieval, content extraction, paper discovery, and claim verification. We evaluate GPT-4o, GPT-5, and Gemini-2.5-Flash under realistic usage conditions-via web interfaces where search operations are opaque to the user. Through controlled experiments, we find consistent reliability failures: citation retrieval fails in 48-98% of multi-reference queries, section-specific content extraction fails in 72-91% of cases, and topical paper discovery yields F1 scores below 0.32, missing over 60% of relevant literature. Further human analysis attributes these failures to the uncontrolled expansion of retrieved context and the tendency of LLMs to prioritize semantically relevant text over task instructions. Across basic tasks, the LLMs display distinct failure behaviors: ChatGPT often withholds responses rather than risk errors, whereas Gemini produces fluent but fabricated answers. To address these issues, we develop lightweight reliability classifiers trained on PaperAsk data to identify unreliable outputs. PaperAsk provides a reproducible and diagnostic framework for advancing the reliability evaluation of LLM-based scholarly assistance systems.