mlsanityData to 14 Oct 2026

Emptying the Ocean with a Spoon: Should We Edit Models?

arXiv 2310.11958 ↗First posted 18 Oct 2023cs.CL cs.LGcited by 27

Not on the map. Why?

Abstract (arXiv)

We call into question the recently popularized method of direct model editing as a means of correcting factual errors in LLM generations. We contrast model editing with three similar but distinct approaches that pursue better defined objectives: (1) retrieval-based architectures, which decouple factual memory from inference and linguistic capabilities embodied in LLMs; (2) concept erasure methods, which aim at preventing systemic bias in generated text; and (3) attribution methods, which aim at grounding generations into identified textual sources. We argue that direct model editing cannot be trusted as a systematic remedy for the disadvantages inherent to LLMs, and while it has proven potential in improving model explainability, it opens risks by reinforcing the notion that models can be trusted for factuality. We call for cautious promotion and application of model editing as part of the LLM deployment process, and for responsibly limiting the use cases of LLMs to those not relying on editing as a critical component.

Key facts. Cited by 27 papers in the map (data through 14 Oct 2026). Builds on 42 of its 54 references that are in the map. First posted on arXiv 18 Oct 2023.

Builds on

42 of 54 references.

References · 42

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2022Training language models to follow instructions with human feedbackcited by 15,971
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  4. 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
  5. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,720
  6. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  7. 2019Language Models as Knowledge Bases?cited by 1,936
  8. 2018Know What You Don't Know: Unanswerable Questions for SQuADcited by 1,554
  9. 2021Improving language models by retrieving from trillions of tokenscited by 1,293
  10. 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,128
  11. 2019How Can We Know What Language Models Know?cited by 959
  12. 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 956
30 more
  1. 2022Atlas: Few-shot Learning with Retrieval Augmented Language Modelscited by 952
  2. 2019Generalization through Memorization: Nearest Neighbor Language Modelscited by 800
  3. 2020How Much Knowledge Can You Pack Into the Parameters of a Language Model?cited by 719
  4. 2021Editing Factual Knowledge in Language Modelscited by 553
  5. 2022Large Language Models Struggle to Learn Long-Tail Knowledgecited by 458
  6. 2020Null It Out: Guarding Protected Attributes by Iterative Nullspace Projectioncited by 455
  7. 2023Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedbackcited by 353
  8. 2023LEACE: Perfect linear concept erasure in closed formcited by 267
  9. 2021Measuring Attribution in Natural Language Generation Modelscited by 256
  10. 2022Is Reinforcement Learning (Not) for Natural Language Processing: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimizationcited by 246
  11. 2019Negated and Misprimed Probes for Pretrained Language Models: Birds Can Talk, But Cannot Flycited by 243
  12. 2023Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Modelscited by 243
  13. 2023Evaluating the Ripple Effects of Knowledge Editing in Language Modelscited by 242
  14. 2020Modifying Memories in Transformer Modelscited by 224
  15. 2022Factuality Enhanced Language Models for Open-Ended Text Generationcited by 202
  16. 2023Progressive Prompts: Continual Learning for Language Modelscited by 197
  17. 2023Grounding Large Language Models in Interactive Environments with Online Reinforcement Learningcited by 195
  18. 2021Towards Continual Knowledge Learning of Language Modelscited by 174
  19. 2018Adversarial Removal of Demographic Attributes from Text Datacited by 170
  20. 2023Head-to-Tail: How Knowledgeable are Large Language Models (LLMs)? A.K.A. Will LLMs Replace Knowledge Graphs?cited by 168
  21. 2022A Review on Language Models as Knowledge Basescited by 157
  22. 2020Editable Neural Networkscited by 155
  23. 2022Training Language Models with Memory Augmentationcited by 113
  24. 2022DisentQA: Disentangling Parametric and Contextual Knowledge with Counterfactual Question Answeringcited by 96
  25. 2020Optimal Continual Learning has Perfect Memory and is NP-hardcited by 87
  26. 2023Can LMs Learn New Entities from Descriptions? Challenges in Propagating Injected Knowledgecited by 67
  27. 2023Detecting Edit Failures In Large Language Models: An Improved Specificity Benchmarkcited by 62
  28. 2023Knowledge Card: Filling LLMs' Knowledge Gaps with Plug-in Specialized Language Modelscited by 46
  29. 2020Facts as Experts: Adaptable and Interpretable Neural Memory over Symbolic Knowledgecited by 37
  30. 2023Language Model Analysis for Ontology Subsumption Inferencecited by 25

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 3; 2024: 16; 2025: 7; 2026: 1.

27 papers.

Sort:

By area:

  1. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,601
  2. 2024Knowledge Conflicts for LLMs: A Surveycited by 283
  3. 2023Knowledge Editing for Large Language Models: A Surveycited by 194
  4. 2024A Comprehensive Study of Knowledge Editing for Large Language Modelscited by 163
  5. 2024Detoxifying Large Language Models via Knowledge Editingcited by 87
  6. 2024Building Guardrails for Large Language Modelscited by 65
  7. 2023Trends in Integration of Knowledge and Large Language Models: A Survey and Taxonomy of Methods, Benchmarks, and Applicationscited by 28
  8. 2024Propagation and Pitfalls: Reasoning-based Assessment of Knowledge Editing through Counterfactual Taskscited by 24
  9. 2024DyKnow: Dynamically Verifying Time-Sensitive Factual Knowledge in LLMscited by 24
  10. 2024Fundamental Problems With Model Editing: How Should Rational Belief Revision Work in LLMs?cited by 24
  11. 2024KnowPO: Knowledge-aware Preference Optimization for Controllable Knowledge Selection in Retrieval-Augmented Language Modelscited by 20
  12. 2024InstructEdit: Instruction-based Knowledge Editing for Large Language Modelscited by 17
15 more
  1. 2024Knowledge Localization: Mission Not Accomplished? Enter Query Localization!cited by 13
  2. 2025CoME: An Unlearning-based Approach to Conflict-free Model Editingcited by 12
  3. 2024Navigating the Dual Facets: A Comprehensive Evaluation of Sequential Memory Editing in Large Language Modelscited by 10
  4. 2025LLMs as Repositories of Factual Knowledge: Limitations and Solutionscited by 10
  5. 2025Mass-Editing Memory with Attention in Transformers: A cross-lingual exploration of knowledgecited by 10
  6. 2024MEMoE: Enhancing Model Editing with Mixture of Experts Adaptorscited by 9
  7. 2025Revealing and Mitigating Over-Attention in Knowledge Editingcited by 8
  8. 2024TAXI: Evaluating Categorical Knowledge Editing for Language Modelscited by 7
  9. 2024Towards Incremental Learning in Large Language Models: A Critical Reviewcited by 6
  10. 2024See the Unseen: Better Context-Consistent Knowledge-Editing by Noisescited by 5
  11. 2025The Knowledge Microscope: Features as Better Analytical Lenses than Neuronscited by 5
  12. 2025Beyond Memorization: A Rigorous Evaluation Framework for Medical Knowledge Editingcited by 4
  13. 2025InComeS: Integrating Compression and Selection Mechanisms into LLMs for Efficient Model Editingcited by 2
  14. 2026Towards Scalable Lifelong Knowledge Editing with Selective Knowledge Suppressioncited by 1
  15. 2024Beyond Individual Facts: Investigating Categorical Knowledge Locality of Taxonomy and Meronomy Concepts in GPT Modelscited by 0