Conflict-Resolving and Sharpness-Aware Minimization for Generalized Knowledge Editing with Multiple Updates

• URL: http://arxiv.org/abs/2602.03696v1
• Date: Tue, 03 Feb 2026 16:18:06 GMT
• Title: Conflict-Resolving and Sharpness-Aware Minimization for Generalized Knowledge Editing with Multiple Updates
• Authors: Duy Nguyen, Hanqi Xiao, Archiki Prasad, Elias Stengel-Eskin, Hyunji Lee, Mohit Bansal,
• Abstract summary: CoRSA is a parameter-efficient, holistic approach for knowledge editing with multiple updates.<n>It tackles multiple challenges simultaneously: it improves generalization to different input forms and enhances stability across multiple updates.<n>CoRSA also generalizes to the code domain, outperforming the strongest baseline by 5.48% Pass@5 in update efficacy.
• Score: 69.6610686845008
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) rely on internal knowledge to solve many downstream tasks, making it crucial to keep them up to date. Since full retraining is expensive, prior work has explored efficient alternatives such as model editing and parameter-efficient fine-tuning. However, these approaches often break down in practice due to poor generalization across inputs, limited stability, and knowledge conflict. To address these limitations, we propose the CoRSA (Conflict-Resolving and Sharpness-Aware Minimization) training framework, a parameter-efficient, holistic approach for knowledge editing with multiple updates. CoRSA tackles multiple challenges simultaneously: it improves generalization to different input forms and enhances stability across multiple updates by minimizing loss curvature, and resolves conflicts by maximizing the margin between new and prior knowledge. Across three widely used fact editing benchmarks, CoRSA achieves significant gains in generalization, outperforming baselines with average absolute improvements of 12.42% over LoRA and 10% over model editing methods. With multiple updates, it maintains high update efficacy while reducing catastrophic forgetting by 27.82% compared to LoRA. CoRSA also generalizes to the code domain, outperforming the strongest baseline by 5.48% Pass@5 in update efficacy.

Related papers

• SA-CAISR: Stage-Adaptive and Conflict-Aware Incremental Sequential Recommendation [34.39526892352457]
  We propose SA-CAISR, a Stage-Adaptive and Conflict-Aware Incremental Sequential Recommendation framework.<n>As a buffer-free framework, SA-CAISR operates using only the old model and new data, directly addressing the high costs of replay-based techniques.<n>We show that SA-CAISR improves Recall@20 by 2.0% on average across datasets, while reducing memory usage by 97.5% and training time by 46.9% compared to the best baseline.
  arXiv  Detail & Related papers  (2026-02-09T14:00:52Z)
• Defeating the Training-Inference Mismatch via FP16 [72.25890308541334]
  Reinforcement learning (RL) fine-tuning often suffers from instability due to the numerical mismatch between the training and inference policies.<n>We show that its root cause lies in the floating point precision itself.<n>The widely adopted BF16, despite its large dynamic range, introduces large rounding errors that breaks the consistency between training and inference.
  arXiv  Detail & Related papers  (2025-10-30T17:58:11Z)
• REPAIR: Robust Editing via Progressive Adaptive Intervention and Reintegration [11.462236606266567]
  Post-training for large language models (LLMs) is constrained by the high cost of acquiring new knowledge or correcting errors.<n>This work introduces a robust framework for developing reliable, scalable, and continually evolving LLMs.
  arXiv  Detail & Related papers  (2025-10-02T10:35:39Z)
• Staying in the Sweet Spot: Responsive Reasoning Evolution via Capability-Adaptive Hint Scaffolding [59.60915947702282]
  Reinforcement learning with verifiable rewards (RLVR) has achieved remarkable success in enhancing the reasoning capabilities of large language models (LLMs)<n>Existing RLVR methods often suffer from exploration inefficiency due to mismatches between the training data's difficulty and the model's capability.<n>We propose SEELE, a novel supervision-aided RLVR framework that dynamically adjusts problem difficulty to stay within the high-efficiency region.
  arXiv  Detail & Related papers  (2025-09-08T17:36:21Z)
• Accelerating Automatic Program Repair with Dual Retrieval-Augmented Fine-Tuning and Patch Generation on Large Language Models [28.75106676284909]
  We propose SelRepair, a novel APR approach with integration of a fine-tuned LLM with a newly-designed dual RAG module.<n>This approach uses a bug-fix pair dataset for fine-tuning and incorporates semantic and syntactic/structural similarity information through an RAG selection gate.<n> Evaluations on Java datasets show SelRepair outperforms other APR methods, achieving 26.29% and 17.64% in terms of exact match (EM) on different datasets while reducing inference time by at least 6.42% with controlled input lengths.
  arXiv  Detail & Related papers  (2025-07-14T09:41:51Z)
• Perception-Aware Policy Optimization for Multimodal Reasoning [79.56070395437898]
  A major source of error in current multimodal reasoning lies in the perception of visual inputs.<n>We propose PAPO, a novel policy gradient algorithm that encourages the model to learn to perceive while learning to reason.<n>We observe a substantial reduction of 30.5% in perception errors, indicating improved perceptual capabilities with PAPO.
  arXiv  Detail & Related papers  (2025-07-08T23:22:34Z)
• Towards Understanding Bugs in Distributed Training and Inference Frameworks for Large Language Models [7.486731499255164]
  This paper conducts the first large-scale empirical analysis of 308 fixed bugs across three popular distributed training/inference frameworks: DeepSpeed, Megatron-LM, and Colossal-AI.<n>We examine bug symptoms, root causes, bug identification and fixing efforts, and common low-effort fixing strategies.
  arXiv  Detail & Related papers  (2025-06-12T07:24:59Z)
• Enhancing Learned Knowledge in LoRA Adapters Through Efficient Contrastive Decoding on Ascend NPUs [12.995753143157083]
  Contrastive LoRA Decoding (CoLD) is a novel decoding framework designed to maximize the use of task-specific knowledge in LoRA-adapted models.<n>CoLD achieves up to a 5.54% increase in task accuracy while reducing end-to-end latency by 28%.
  arXiv  Detail & Related papers  (2025-05-20T17:11:18Z)
• LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging [80.17238673443127]
  LiNeS is a post-training editing technique designed to preserve pre-trained generalization while enhancing fine-tuned task performance.<n>LiNeS demonstrates significant improvements in both single-task and multi-task settings across various benchmarks in vision and natural language processing.
  arXiv  Detail & Related papers  (2024-10-22T16:26:05Z)
• O-Edit: Orthogonal Subspace Editing for Language Model Sequential Editing [0.0]
  Large language models (LLMs) acquire knowledge during pre-training, but over time, this knowledge may become incorrect or outdated, necessitating updates after training. We propose Orthogonal Subspace Editing, O-Edit. This algorithmizes the direction of each knowledge update, minimizing interference between successive updates and reducing the impact of new updates on unrelated knowledge. It can perform thousands of edits on mainstream LLMs, achieving an average performance improvement that is 4.2 times better than existing methods while effectively preserving the model's performance on downstream tasks, all with minimal additional parameter overhead.
  arXiv  Detail & Related papers  (2024-10-15T10:16:45Z)
• Masked Thought: Simply Masking Partial Reasoning Steps Can Improve Mathematical Reasoning Learning of Language Models [102.72940700598055]
  In reasoning tasks, even a minor error can cascade into inaccurate results. We develop a method that avoids introducing external resources, relying instead on perturbations to the input. Our training approach randomly masks certain tokens within the chain of thought, a technique we found to be particularly effective for reasoning tasks.
  arXiv  Detail & Related papers  (2024-03-04T16:21:54Z)
• Defensive Patches for Robust Recognition in the Physical World [111.46724655123813]
  Data-end defense improves robustness by operations on input data instead of modifying models. Previous data-end defenses show low generalization against diverse noises and weak transferability across multiple models. We propose a defensive patch generation framework to address these problems by helping models better exploit these features.
  arXiv  Detail & Related papers  (2022-04-13T07:34:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.