# Do LLMs Know When to NOT Answer? Investigating Abstention Abilities of Large Language Models

arXiv: https://arxiv.org/abs/2407.16221 · HTML: https://mlsanity.com/paper/2407.16221/

## Key facts

- Cited by 59 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 41 references that are in the map.
- First posted on arXiv 23 Jul 2024.

## Builds on (most cited first, 26 of 26)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories](https://mlsanity.com/paper/2212.10511/) (2022)
- [Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMs](https://mlsanity.com/paper/2306.13063/) (2023)
- [Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback](https://mlsanity.com/paper/2305.14975/) (2023)
- [Crowdsourcing Multiple Choice Science Questions](https://mlsanity.com/paper/1707.06209/) (2017)
- [The Internal State of an LLM Knows When It's Lying](https://mlsanity.com/paper/2304.13734/) (2023)
- [UnifiedQA: Crossing Format Boundaries With a Single QA System](https://mlsanity.com/paper/2005.00700/) (2020)
- [Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs](https://mlsanity.com/paper/2406.15927/) (2024)
- [Large Language Models in Finance: A Survey](https://mlsanity.com/paper/2311.10723/) (2023)
- [Large Language Models Understand and Can be Enhanced by Emotional Stimuli](https://mlsanity.com/paper/2307.11760/) (2023)
- [Quantifying Aleatoric and Epistemic Uncertainty in Machine Learning: Are Conditional Entropy and Mutual Information Appropriate Measures?](https://mlsanity.com/paper/2209.03302/) (2022)
- [SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales](https://mlsanity.com/paper/2405.20974/) (2024)
- [Conformal Alignment: Knowing When to Trust Foundation Models with Guarantees](https://mlsanity.com/paper/2405.10301/) (2024)
- [Uncertainty-Based Abstention in LLMs Improves Safety and Reduces Hallucinations](https://mlsanity.com/paper/2404.10960/) (2024)
- [From Sparse to Dense: GPT-4 Summarization with Chain of Density Prompting](https://mlsanity.com/paper/2309.04269/) (2023)
- [Uncertainty quantification in fine-tuned LLMs using LoRA ensembles](https://mlsanity.com/paper/2402.12264/) (2024)
- [Exploring Predictive Uncertainty and Calibration in NLP: A Study on the Impact of Method & Data Scarcity](https://mlsanity.com/paper/2210.15452/) (2022)
- [Post-Abstention: Towards Reliably Re-Attempting the Abstained Instances in QA](https://mlsanity.com/paper/2305.01812/) (2023)
- [DeepSRGM -- Sequence Classification and Ranking in Indian Classical Music with Deep Learning](https://mlsanity.com/paper/2402.10168/) (2024)

## Built on by (most cited first, 30 of 59)

- [AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions](https://mlsanity.com/paper/2506.09038/) (2025)
- [AgentAbstain: Do LLM Agents Know When Not to Act?](https://mlsanity.com/paper/2607.10059/) (2026)
- [Knowing When to Abstain: Medical LLMs Under Clinical Uncertainty](https://mlsanity.com/paper/2601.12471/) (2026)
- [Whose Facts Win? LLM Source Preferences under Knowledge Conflicts](https://mlsanity.com/paper/2601.03746/) (2026)
- [Wait, that's not an option: LLMs Robustness with Incorrect Multiple-Choice Options](https://mlsanity.com/paper/2409.00113/) (2024)
- [Cannot or Should Not? Automatic Analysis of Refusal Composition in IFT/RLHF Datasets and Refusal Behavior of Black-Box LLMs](https://mlsanity.com/paper/2412.16974/) (2024)
- [Do not Abstain! Identify and Solve the Uncertainty](https://mlsanity.com/paper/2506.00780/) (2025)
- [Over-Searching in Search-Augmented Large Language Models](https://mlsanity.com/paper/2601.05503/) (2026)
- [CausalAbstain: Enhancing Multilingual LLMs with Causal Reasoning for Trustworthy Abstention](https://mlsanity.com/paper/2506.00519/) (2025)
- [Causal Evidence that Language Models use Confidence to Drive Behavior](https://mlsanity.com/paper/2603.22161/) (2026)
- [RealFin: How Well Do LLMs Reason About Finance When Users Leave Things Unsaid?](https://mlsanity.com/paper/2602.07096/) (2026)
- [Mitigating Manipulation and Enhancing Persuasion: A Reflective Multi-Agent Approach for Legal Argument Generation](https://mlsanity.com/paper/2506.02992/) (2025)
- [Stop Before You Fail: Operational Capability Boundaries for Mitigating Unproductive Reasoning in Large Reasoning Models](https://mlsanity.com/paper/2509.24711/) (2025)
- [Can LLMs Estimate Student Struggles? Human-AI Difficulty Alignment with Proficiency Simulation for Item Difficulty Prediction](https://mlsanity.com/paper/2512.18880/) (2025)
- [BAS: A Decision-Theoretic Approach to Evaluating Large Language Model Confidence](https://mlsanity.com/paper/2604.03216/) (2026)
- [Who's Asking? Investigating Bias Through the Lens of Disability Framed Queries in LLMs](https://mlsanity.com/paper/2508.15831/) (2025)
- [Mitigating Lost in Multi-turn Conversation via Curriculum RL with Verifiable Accuracy and Abstention Rewards](https://mlsanity.com/paper/2510.18731/) (2025)
- [Comparing Uncertainty Measurement and Mitigation Methods for Large Language Models: A Systematic Review](https://mlsanity.com/paper/2504.18346/) (2025)
- [FESTA: Functionally Equivalent Sampling for Trust Assessment of Multimodal LLMs](https://mlsanity.com/paper/2509.16648/) (2025)
- [Honesty over Accuracy: Trustworthy Language Models through Reinforced Hesitation](https://mlsanity.com/paper/2511.11500/) (2025)
- [Variational Uncertainty Decomposition for In-Context Learning](https://mlsanity.com/paper/2509.02327/) (2025)
- [ClinDet-Bench: Beyond Abstention, Evaluating Judgment Determinability of LLMs in Clinical Decision-Making](https://mlsanity.com/paper/2602.22771/) (2026)
- [The Yes-Man Syndrome: Benchmarking Abstention in Embodied Robotic Agents](https://mlsanity.com/paper/2605.20544/) (2026)
- [Towards Harmonized Uncertainty Estimation for Large Language Models](https://mlsanity.com/paper/2505.19073/) (2025)
- [Knowing What's Missing: Assessing Information Sufficiency in Question Answering](https://mlsanity.com/paper/2512.06476/) (2025)
- [Know Or Not: a library for evaluating out-of-knowledge base robustness](https://mlsanity.com/paper/2505.13545/) (2025)
- [Who Gets the Kidney? Human-AI Alignment, Indecision, and Moral Values](https://mlsanity.com/paper/2506.00079/) (2025)
- [Scaling Truth: The Confidence Paradox in AI Fact-Checking](https://mlsanity.com/paper/2509.08803/) (2025)
- [Partial Evidence Bench: Benchmarking Authorization-Limited Evidence in Agentic Systems](https://mlsanity.com/paper/2605.05379/) (2026)
- [Large Language Models Lack Temporal Awareness of Medical Knowledge](https://mlsanity.com/paper/2605.13045/) (2026)

## Abstract (arXiv)

Abstention Ability (AA) is a critical aspect of Large Language Model (LLM) reliability, referring to an LLM's capability to withhold responses when uncertain or lacking a definitive answer, without compromising performance. Although previous studies have attempted to improve AA, they lack a standardised evaluation method and remain unsuitable for black-box models where token prediction probabilities are inaccessible. This makes comparative analysis challenging, especially for state-of-the-art closed-source commercial LLMs. This paper bridges this gap by introducing a black-box evaluation approach and a new dataset, Abstain-QA, crafted to rigorously assess AA across varied question types (answerable and unanswerable), domains (well-represented and under-represented), and task types (fact centric and reasoning). We also propose a new confusion matrix, the ''Answerable-Unanswerable Confusion Matrix'' (AUCM) which serves as the basis for evaluating AA, by offering a structured and precise approach for assessment. Finally, we explore the impact of three prompting strategies-Strict Prompting, Verbal Confidence Thresholding, and Chain-of-Thought (CoT)-on improving AA. Our results indicate that even powerful models like GPT-4, Mixtral 8x22b encounter difficulties with abstention; however, strategic approaches such as Strict prompting and CoT can enhance this capability.