mlsanityData to 7 Oct 2026

Composable Interventions for Language Models

arXiv 2407.06483 ↗First posted 9 Jul 2024cs.LG cs.CLcited by 4

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

Test-time interventions for language models can enhance factual accuracy, mitigate harmful outputs, and improve model efficiency without costly retraining. But despite a flood of new methods, different types of interventions are largely developing independently. In practice, multiple interventions must be applied sequentially to the same model, yet we lack standardized ways to study how interventions interact. We fill this gap by introducing composable interventions, a framework to study the effects of using multiple interventions on the same language models, featuring new metrics and a unified codebase. Using our framework, we conduct extensive experiments and compose popular methods from three emerging intervention categories -- Knowledge Editing, Model Compression, and Machine Unlearning. Our results from 310 different compositions uncover meaningful interactions: compression hinders editing and unlearning, composing interventions hinges on their order of application, and popular general-purpose metrics are inadequate for assessing composability. Taken together, our findings showcase clear gaps in composability, suggesting a need for new multi-objective interventions. All of our code is public: https://github.com/hartvigsen-group/composable-interventions.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 88 of its 106 references that are in the map. First posted on arXiv 9 Jul 2024.

Builds on

88 of 106 references.

Closest ancestors · 10

Same direction, newest first

  1. 2024Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning8 Apr
  2. 2024Eight Methods to Evaluate Robust Unlearning in LLMs26 Feb
  3. 2024Machine Unlearning of Pre-trained Large Language Models23 Feb
  4. 2024Towards Safer Large Language Models through Machine Unlearning15 Feb
  5. 2024Rethinking Machine Unlearning for Large Language Models13 Feb
  6. 2024TOFU: A Task of Fictitious Unlearning for LLMs11 Jan
  7. 2023Who's Harry Potter? Approximate Unlearning in LLMs3 Oct
  8. 2023Exploring the Landscape of Machine Unlearning: A Comprehensive Survey and Taxonomy10 May
  9. 2023Model Sparsity Can Simplify Machine Unlearning11 Apr
  10. 2022Knowledge Unlearning for Mitigating Privacy Risks in Language Models4 Oct

Other references · 78

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  4. 2016Pointer Sentinel Mixture Modelscited by 2,992
  5. 2022Locating and Editing Factual Associations in GPTcited by 2,609
  6. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  7. 2022GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformerscited by 1,838
  8. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,338
  9. 2023Representation Engineering: A Top-Down Approach to AI Transparencycited by 1,239
  10. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,235
  11. 2023SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shotcited by 1,090
  12. 2022Mass-Editing Memory in a Transformercited by 976
38 more
  1. 2023BloombergGPT: A Large Language Model for Financecited by 923
  2. 2022LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scalecited by 881
  3. 2018The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networkscited by 847
  4. 2023A Simple and Effective Pruning Approach for Large Language Modelscited by 776
  5. 2023Large Language Modelscited by 774
  6. 2022Galactica: A Large Language Model for Sciencecited by 736
  7. 2023Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizescited by 710
  8. 2024Yi: Open Foundation Models by 01.AIcited by 706
  9. 2023LLM-Pruner: On the Structural Pruning of Large Language Modelscited by 643
  10. 2019Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networkscited by 599
  11. 2021Editing Factual Knowledge in Language Modelscited by 548
  12. 2019The Woman Worked as a Babysitter: On Biases in Language Generationcited by 540
  13. 2022ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformerscited by 517
  14. 2021Fast Model Editing at Scalecited by 491
  15. 2024The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearningcited by 488
  16. 2022Memory-Based Model Editing at Scalecited by 443
  17. 2017Zero-Shot Relation Extraction via Reading Comprehensioncited by 425
  18. 2022ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detectioncited by 419
  19. 2023Towards Expert-Level Medical Question Answering with Large Language Modelscited by 382
  20. 2023Editing Large Language Models: Problems, Methods, and Opportunitiescited by 367
  21. 2023MQuAKE: Assessing Knowledge Editing in Language Models via Multi-Hop Questionscited by 292
  22. 2022Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptorscited by 267
  23. 2023A Survey on Model Compression for Large Language Modelscited by 263
  24. 2023Challenges and Applications of Large Language Modelscited by 251
  25. 2023Evaluating the Ripple Effects of Knowledge Editing in Language Modelscited by 238
  26. 2023Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Modelscited by 237
  27. 2023Transformer-Patcher: One Mistake worth One Neuroncited by 196
  28. 2023LoftQ: LoRA-Fine-Tuning-Aware Quantization for Large Language Modelscited by 195
  29. 2023Knowledge Editing for Large Language Models: A Surveycited by 189
  30. 2024Lessons from the Trenches on Reproducible Evaluation of Language Modelscited by 172
  31. 2021Challenges in Detoxifying Language Modelscited by 169
  32. 2019What Do Compressed Deep Neural Networks Forget?cited by 164
  33. 2023Outlier Weighed Layerwise Sparsity (OWL): A Missing Secret Sauce for Pruning LLMs to High Sparsitycited by 163
  34. 2020Editable Neural Networkscited by 154
  35. 2023Emergent and Predictable Memorization in Large Language Modelscited by 147
  36. 2021The Benchmark Lotterycited by 133
  37. 2023The Truth is in There: Improving Reasoning in Language Models with Layer-Selective Rank Reductioncited by 120
  38. 2023Copyright Violations and Large Language Modelscited by 108

50 most cited of 78 All 78 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 3; 2025: 0; 2026: 1.

4 papers.

  1. 2024Steering Language Model Refusal with Sparse Autoencoderscited by 89
  2. 2024Catastrophic Failure of LLM Unlearning via Quantizationcited by 76
  3. 2024Position: LLM Unlearning Benchmarks are Weak Measures of Progresscited by 58
  4. 2026Activation-Space Order-Swap Geometry: A Site-Asymmetry Auditcited by 0