# MEMOIR: Lifelong Model Editing with Minimal Overwrite and Informed Retention for LLMs

arXiv: https://arxiv.org/abs/2506.07899 · HTML: https://mlsanity.com/paper/2506.07899/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 32 of its 44 references that are in the map.
- First posted on arXiv 9 Jun 2025.

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

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://mlsanity.com/paper/2101.00027/) (2020)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [Transformer Feed-Forward Layers Are Key-Value Memories](https://mlsanity.com/paper/2012.14913/) (2020)
- [Mass-Editing Memory in a Transformer](https://mlsanity.com/paper/2210.07229/) (2022)
- [PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning](https://mlsanity.com/paper/1711.05769/) (2017)
- [SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models](https://mlsanity.com/paper/2303.08896/) (2023)
- [An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks](https://mlsanity.com/paper/1312.6211/) (2013)
- [Fast Model Editing at Scale](https://mlsanity.com/paper/2110.11309/) (2021)
- [Memory-Based Model Editing at Scale](https://mlsanity.com/paper/2206.06520/) (2022)
- [Zero-Shot Relation Extraction via Reading Comprehension](https://mlsanity.com/paper/1706.04115/) (2017)
- [Gradient Projection Memory for Continual Learning](https://mlsanity.com/paper/2103.09762/) (2021)
- [Can We Edit Factual Knowledge by In-Context Learning?](https://mlsanity.com/paper/2305.12740/) (2023)
- [Continual learning with hypernetworks](https://mlsanity.com/paper/1906.00695/) (2019)
- [Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptors](https://mlsanity.com/paper/2211.11031/) (2022)
- [Evaluating the Ripple Effects of Knowledge Editing in Language Models](https://mlsanity.com/paper/2307.12976/) (2023)
- [Supermasks in Superposition](https://mlsanity.com/paper/2006.14769/) (2020)
- [Transformer-Patcher: One Mistake worth One Neuron](https://mlsanity.com/paper/2301.09785/) (2023)
- [Understanding the Role of Training Regimes in Continual Learning](https://mlsanity.com/paper/2006.06958/) (2020)
- [Should ChatGPT be Biased? Challenges and Risks of Bias in Large Language Models](https://mlsanity.com/paper/2304.03738/) (2023)
- [Distributionally Robust Language Modeling](https://mlsanity.com/paper/1909.02060/) (2019)
- [EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Models](https://mlsanity.com/paper/2308.07269/) (2023)
- [WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models](https://mlsanity.com/paper/2405.14768/) (2024)
- [SparCL: Sparse Continual Learning on the Edge](https://mlsanity.com/paper/2209.09476/) (2022)
- [Correcting Diverse Factual Errors in Abstractive Summarization via Post-Editing and Language Model Infilling](https://mlsanity.com/paper/2210.12378/) (2022)

## Built on by (most cited first, 10 of 10)

- [EvoEdit: Evolving Null-space Alignment for Robust and Efficient Knowledge Editing](https://mlsanity.com/paper/2510.13851/) (2025)
- [Label-Free Cross-Task LoRA Merging with Null-Space Compression](https://mlsanity.com/paper/2603.26317/) (2026)
- [Representation Interventions Enable Lifelong Knowledge Memory Control in LLMs](https://mlsanity.com/paper/2511.20892/) (2025)
- [Preference-Aligned LoRA Merging: Preserving Subspace Coverage and Addressing Directional Anisotropy](https://mlsanity.com/paper/2603.26299/) (2026)
- [Norm Anchors Make Model Edits Last](https://mlsanity.com/paper/2602.02543/) (2026)
- [Interactionless Inverse Reinforcement Learning: A Data-Centric Framework for Durable Alignment](https://mlsanity.com/paper/2602.14844/) (2026)
- [Soul Computing: A Theoretical Framework and Technical Architecture for Intelligent Agents with Independent Consciousness](https://mlsanity.com/paper/2606.10413/) (2026)
- [Beyond Endpoint Scores: Time- and Capacity-Conditioned Evaluation of Continual Knowledge Updating](https://mlsanity.com/paper/2609.03900/) (2026)
- [ManiEdit: Sequential Unstructured Knowledge Editing for Language Models from a Manifold Perspective](https://mlsanity.com/paper/2609.33534/) (2026)
- [MedKIT: Evaluating Knowledge Integration and Generalization in Large Language Models](https://mlsanity.com/paper/2609.38543/) (2026)

## Abstract (arXiv)

Language models deployed in real-world systems often require post-hoc updates to incorporate new or corrected knowledge. However, editing such models efficiently and reliably-without retraining or forgetting previous information-remains a major challenge. Existing methods for lifelong model editing either compromise generalization, interfere with past edits, or fail to scale to long editing sequences. We propose MEMOIR, a novel scalable framework that injects knowledge through a residual memory, i.e., a dedicated parameter module, while preserving the core capabilities of the pre-trained model. By sparsifying input activations through sample-dependent masks, MEMOIR confines each edit to a distinct subset of the memory parameters, minimizing interference among edits. At inference, it identifies relevant edits by comparing the sparse activation patterns of new queries to those stored during editing. This enables generalization to rephrased queries by activating only the relevant knowledge while suppressing unnecessary memory activation for unrelated prompts. Experiments on question answering, hallucination correction, and out-of-distribution generalization benchmarks for LLaMA-3 and Mistral backbones demonstrate that MEMOIR achieves state-of-the-art performance across reliability, generalization, and locality metrics, scaling to thousands of sequential edits with minimal forgetting.