# Discovering Knowledge Deficiencies of Language Models on Massive Knowledge Base

arXiv: https://arxiv.org/abs/2503.23361 · HTML: https://mlsanity.com/paper/2503.23361/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 32 of its 41 references that are in the map.
- First posted on arXiv 30 Mar 2025.

## Builds on (most cited first, 30 of 32)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [Language Models as Knowledge Bases?](https://mlsanity.com/paper/1909.01066/) (2019)
- [How Can We Know What Language Models Know?](https://mlsanity.com/paper/1911.12543/) (2019)
- [SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models](https://mlsanity.com/paper/2303.08896/) (2023)
- [Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models](https://mlsanity.com/paper/2309.01219/) (2023)
- [Humanity's Last Exam](https://mlsanity.com/paper/2501.14249/) (2025)
- [Dynabench: Rethinking Benchmarking in NLP](https://mlsanity.com/paper/2104.14337/) (2021)
- [SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models](https://mlsanity.com/paper/2307.10635/) (2023)
- [mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval](https://mlsanity.com/paper/2407.19669/) (2024)
- [DyVal: Dynamic Evaluation of Large Language Models for Reasoning Tasks](https://mlsanity.com/paper/2309.17167/) (2023)
- [TrueTeacher: Learning Factual Consistency Evaluation with Large Language Models](https://mlsanity.com/paper/2305.11171/) (2023)
- [Benchmark Self-Evolving: A Multi-Agent Framework for Dynamic LLM Evaluation](https://mlsanity.com/paper/2402.11443/) (2024)
- [ConflictBank: A Benchmark for Evaluating the Influence of Knowledge Conflicts in LLM](https://mlsanity.com/paper/2408.12076/) (2024)
- [Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering](https://mlsanity.com/paper/2410.15999/) (2024)
- [On Large Language Models' Hallucination with Regard to Known Facts](https://mlsanity.com/paper/2403.20009/) (2024)
- [On the Trustworthiness of Generative Foundation Models: Guideline, Assessment, and Perspective](https://mlsanity.com/paper/2502.14296/) (2025)
- [Mechanistic Understanding and Mitigation of Language Model Non-Factual Hallucinations](https://mlsanity.com/paper/2403.18167/) (2024)
- [Task Me Anything](https://mlsanity.com/paper/2406.11775/) (2024)
- [Safer-Instruct: Aligning Language Models with Automated Preference Data](https://mlsanity.com/paper/2311.08685/) (2023)
- [EvalTree: Profiling Language Model Weaknesses via Hierarchical Capability Trees](https://mlsanity.com/paper/2503.08893/) (2025)
- [AD-LLM: Benchmarking Large Language Models for Anomaly Detection](https://mlsanity.com/paper/2412.11142/) (2024)
- [DataGen: Unified Synthetic Dataset Generation via Large Language Models](https://mlsanity.com/paper/2406.18966/) (2024)
- [ProVision: Programmatically Scaling Vision-centric Instruction Data for Multimodal Language Models](https://mlsanity.com/paper/2412.07012/) (2024)
- [NLP-ADBench: NLP Anomaly Detection Benchmark](https://mlsanity.com/paper/2412.04784/) (2024)
- [Distinguishing Ignorance from Error in LLM Hallucinations](https://mlsanity.com/paper/2410.22071/) (2024)
- [Automated Capability Discovery via Foundation Model Self-Exploration](https://mlsanity.com/paper/2502.07577/) (2025)

## Built on by (most cited first, 6 of 6)

- [ProbeLLM: Automating Principled Diagnosis of LLM Failures](https://mlsanity.com/paper/2602.12966/) (2026)
- [Shifting Uncertainty to Critical Moments: Towards Reliable Uncertainty Quantification for VLA Model](https://mlsanity.com/paper/2603.18342/) (2026)
- [Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in LLMs](https://mlsanity.com/paper/2505.07184/) (2025)
- [M3OOD: Automatic Selection of Multimodal OOD Detectors](https://mlsanity.com/paper/2508.11936/) (2025)
- [ReasoningLens: Hierarchical Visualization and Diagnostic Auditing for Large Reasoning Models](https://mlsanity.com/paper/2606.23404/) (2026)
- [A Graph Perspective to Probe Structural Patterns of Knowledge in Large Language Models](https://mlsanity.com/paper/2505.19286/) (2025)

## Abstract (arXiv)

Large language models (LLMs) possess impressive linguistic capabilities but often fail to faithfully retain factual knowledge, leading to hallucinations and unreliable outputs. Understanding LLMs' knowledge deficiencies by exhaustively evaluating against full-scale knowledge bases is computationally prohibitive, especially for closed-weight models. We propose stochastic error ascent (SEA), a scalable and efficient framework for discovering knowledge deficiencies (errors) in closed-weight LLMs under a strict query budget. Rather than naively probing all knowledge candidates, SEA formulates error discovery as a stochastic optimization process: it iteratively retrieves new high-error candidates by leveraging the semantic similarity to previously observed failures. To further enhance search efficiency and coverage, SEA employs hierarchical retrieval across document and paragraph levels, and constructs a relation directed acyclic graph to model error propagation and identify systematic failure modes. Empirically, SEA uncovers 40.7x more knowledge errors than Automated Capability Discovery and 26.7% more than AutoBencher, while reducing the cost-per-error by 599x and 9x, respectively. Human evaluation confirms the high quality of generated questions, while ablation and convergence analyses validate the contribution of each component in SEA. Further analysis on the discovered errors reveals correlated failure patterns across LLM families and recurring deficits, highlighting the need for better data coverage and targeted fine-tuning in future LLM development.