Tag
该论文研究在数据筛选(如去重)之后实现精确机器遗忘的问题,提出了一种'保留契约',说明服务需要保留哪些信息才能在删除请求后精确复现训练过程,并在AskUbuntu数据集上验证了精确成员资格修复的有效性。
提出 PFDF，一种基于路径依赖的联邦遗忘客户端选择框架，通过 CLP 加权历史记录筛选出与目标客户端弱耦合但全局信息量大的保留客户端，以更少的通信和计算成本实现高效遗忘。
H2DF introduces a calibrated target-band objective for language-model unlearning that limits gradient-ascent pressure once forget loss reaches a data-dependent target range, matching GradDiff performance on TOFU while using far less retain replay and improving retention over GA and NPO on MUSE-News.
论文提出正交边界遗忘（OBU）框架，通过在激活协方差的正交零空间中限制参数更新，在大模型中消除双用领域（如网络安全、生物、化学、核工程）的危险能力，同时不损害良性科学能力，避免概念混淆。
论文提出 CoLU（Confidence-Weighted Layerwise Unlearning），一种针对大型语言模型的置信度加权逐层遗忘方法，并在 Phi-1.5 模型的所有层上报告了完整数据的实验结果。
该论文提出 Forget-Aware Self-Distillation (FASD) 方法，将多模态大模型遗忘解耦为行为通道与知识通道，以实现「诚实无知」——既真正擦除目标知识，又让模型能承认自身知识边界，避免幻觉与信息泄露。
REDaCT提出一种面向大型推理模型的推理时无遗忘框架，通过二元泄漏检测器与纠正性嵌入优化，在彻底移除遗忘集信息的同时完全避免推理链崩溃，实现近乎完美的答案与CoT遗忘并显著保留推理能力。
The paper introduces ExecCert, a release-time certification layer that bridges the gap between mathematical certified machine-unlearning guarantees and the concrete finite-precision artifacts actually deployed, including an incremental Retraining-Reference Release Verification for sequential deletion requests.
The paper introduces SSU-LSF, the first machine-unlearning framework for Mamba-family structured state space models applied to land surface forecasting, using EKFac influence functions and projected gradient ascent to remove non-stationary confounding bias from NDVI, LST, and crop phenology predictions at 8.4x lower cost than full retraining.
The paper introduces Inverse Distillation Unlearning (IDU), a unified framework that simultaneously distills a multi-step flow-matching or diffusion teacher into an efficient one-step student while suppressing generation of forgotten training data, requiring only the teacher and forget-set samples without access to retained data or auxiliary classifiers.
The paper introduces target-speaker unlearning ASR (TSU-ASR) and proposes a novel Enrollment-Conditioned Gating module for dynamic opt-out of speakers during inference in LLM-based ASR, enhancing privacy in online conferencing.
The paper identifies sequential reappearance as a failure mode in diffusion data-point unlearning and proposes a sharpness-guided method to improve forgetting persistence across deletion sequences.
The paper introduces GUARD, a method for natural forgetting in large reasoning models that uses guided answer-reasoning distillation to suppress unsafe or private content in chain-of-thought traces while preserving reasoning utility.
This paper introduces μ²-Bench, a benchmark for evaluating multilingual machine unlearning in large language models, aiming to ensure that undesired information is effectively removed across diverse languages.
ARIA is a test-time unlearning method for large language models that uses sparse autoencoders to suppress unwanted knowledge during inference without modifying weights, improving the forget-retain trade-off and remaining robust to adversarial attacks.
This paper explores machine unlearning techniques for Large Audio-Language Models to remove sensitive information from speech QA tasks, demonstrating methods that reduce privacy leakage by up to 80% while maintaining performance.
The paper introduces forget-set misalignment in LLM unlearning and proposes a data-blind framework called CONFS to address it, achieving a competitive forgetting-utility balance.
This paper introduces I-CARE, a methodology for systematically analyzing interference in machine unlearning for text-to-image models, providing formal definitions and an open-source framework to enable reproducible study.
This paper proposes CallosumNet, a biologically inspired framework for efficient unlearning in spatio-temporal graphs to comply with privacy regulations like GDPR, achieving complete unlearning with minimal accuracy loss.
IsleNet introduces a spatial-entropy-based partitioning method for spatiotemporal graph unlearning, enabling exact data removal with low computational cost while maintaining high accuracy for privacy regulations.