A Comprehensive Evaluation of LLM Unlearning Robustness under Multi-Turn Interaction

• URL: http://arxiv.org/abs/2603.00823v1
• Date: Sat, 28 Feb 2026 22:01:31 GMT
• Title: A Comprehensive Evaluation of LLM Unlearning Robustness under Multi-Turn Interaction
• Authors: Ruihao Pan, Suhang Wang,
• Abstract summary: We study whether unlearning remains stable in interactive environments by examining two common interaction patterns.<n>We find that knowledge appearing forgotten in static evaluation can often be recovered through interaction.
• Score: 28.143025731350352
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine unlearning aims to remove the influence of specific training data from pre-trained models without retraining from scratch, and is increasingly important for large language models (LLMs) due to safety, privacy, and legal concerns. Although prior work primarily evaluates unlearning in static, single-turn settings, forgetting robustness under realistic interactive use remains underexplored. In this paper, we study whether unlearning remains stable in interactive environments by examining two common interaction patterns: self-correction and dialogue-conditioned querying. We find that knowledge appearing forgotten in static evaluation can often be recovered through interaction. Although stronger unlearning improves apparent robustness, it often results in behavioral rigidity rather than genuine knowledge erasure. Our findings suggest that static evaluation may overestimate real-world effectiveness and highlight the need for ensuring stable forgetting under interactive settings.

Related papers

• Auditing Language Model Unlearning via Information Decomposition [68.48660428111593]
  We introduce an interpretable, information-theoretic framework for auditing unlearning using Partial Information Decomposition (PID)<n>By comparing model representations before and after unlearning, we decompose the mutual information with the forgotten data into distinct components, formalizing the notions of unlearned and residual knowledge.<n>Our work introduces a principled, representation-level audit for unlearning, offering theoretical insight and actionable tools for safer deployment of language models.
  arXiv  Detail & Related papers  (2026-01-21T15:51:19Z)
• REMIND: Input Loss Landscapes Reveal Residual Memorization in Post-Unlearning LLMs [0.1784233255402269]
  Machine unlearning aims to remove the influence of specific training data from a model without requiring full retraining.<n>We propose REMIND, a novel evaluation method aiming to detect the subtle remaining influence of unlearned data.<n>We show that unlearned data yield flatter, less steep loss landscapes, while retained or unrelated data exhibit sharper, more volatile patterns.
  arXiv  Detail & Related papers  (2025-11-06T09:58:19Z)
• Edu-EmotionNet: Cross-Modality Attention Alignment with Temporal Feedback Loops [0.0]
  We introduce Edu-EmotionNet, a novel framework that jointly models temporal emotion evolution and modality reliability for robust affect recognition.<n>EmotionNet achieves state-of-the-art performance and demonstrates strong robustness to missing or noisy modalities.
  arXiv  Detail & Related papers  (2025-10-09T20:33:52Z)
• Existing Large Language Model Unlearning Evaluations Are Inconclusive [105.55899615056573]
  We show that some evaluations introduce substantial new information into the model, potentially masking true unlearning performance.<n>We demonstrate that evaluation outcomes vary significantly across tasks, undermining the generalizability of current evaluation routines.<n>We propose two principles for future unlearning evaluations: minimal information injection and downstream task awareness.
  arXiv  Detail & Related papers  (2025-05-31T19:43:00Z)
• Reinforced Interactive Continual Learning via Real-time Noisy Human Feedback [59.768119380109084]
  This paper introduces an interactive continual learning paradigm where AI models dynamically learn new skills from real-time human feedback.<n>We propose RiCL, a Reinforced interactive Continual Learning framework leveraging Large Language Models (LLMs)<n>Our RiCL approach substantially outperforms existing combinations of state-of-the-art online continual learning and noisy-label learning methods.
  arXiv  Detail & Related papers  (2025-05-15T03:22:03Z)
• Robustness Evaluation of Offline Reinforcement Learning for Robot Control Against Action Perturbations [4.849820402342814]
  offline reinforcement learning is particularly promising for robot control applications.<n> robustness against real-world challenges, such as joint actuator faults in robots, remains a critical concern.<n>This study evaluates the robustness of existing offline reinforcement learning methods using legged robots from OpenAI Gym.
  arXiv  Detail & Related papers  (2024-12-25T05:02:22Z)
• Towards Effective Evaluations and Comparisons for LLM Unlearning Methods [97.2995389188179]
  This paper seeks to refine the evaluation of machine unlearning for large language models.<n>It addresses two key challenges -- the robustness of evaluation metrics and the trade-offs between competing goals.
  arXiv  Detail & Related papers  (2024-06-13T14:41:00Z)
• AntEval: Evaluation of Social Interaction Competencies in LLM-Driven Agents [65.16893197330589]
  Large Language Models (LLMs) have demonstrated their ability to replicate human behaviors across a wide range of scenarios. However, their capability in handling complex, multi-character social interactions has yet to be fully explored. We introduce the Multi-Agent Interaction Evaluation Framework (AntEval), encompassing a novel interaction framework and evaluation methods.
  arXiv  Detail & Related papers  (2024-01-12T11:18:00Z)
• Stateful Offline Contextual Policy Evaluation and Learning [88.9134799076718]
  We study off-policy evaluation and learning from sequential data. We formalize the relevant causal structure of problems such as dynamic personalized pricing. We show improved out-of-sample policy performance in this class of relevant problems.
  arXiv  Detail & Related papers  (2021-10-19T16:15:56Z)
• Towards Automatic Evaluation of Dialog Systems: A Model-Free Off-Policy Evaluation Approach [84.02388020258141]
  We propose a new framework named ENIGMA for estimating human evaluation scores based on off-policy evaluation in reinforcement learning. ENIGMA only requires a handful of pre-collected experience data, and therefore does not involve human interaction with the target policy during the evaluation. Our experiments show that ENIGMA significantly outperforms existing methods in terms of correlation with human evaluation scores.
  arXiv  Detail & Related papers  (2021-02-20T03:29:20Z)

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.