NPO enables stable unlearning of 50%+ training data in LLMs on TOFU by making collapse exponentially slower than gradient ascent, preserving sensible outputs where prior methods fail.
Varshney, Mohit Bansal, Sanmi Koyejo, and Yang Liu
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
representative citing papers
TRACER uses token reassignment for concept-related items plus a coherence regularizer to unlearn specific concepts in generative recommendation while preserving utility better than baselines.
Unlearning objectives should be tailored to distinct language functions, with a meta-learned RMU variant for dangerous knowledge and a multi-layer probe objective for toxicity, yielding strong results on four 7-8B models.
REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.
CatShift detects training data membership in LLMs by comparing output shifts induced by fine-tuning on member versus non-member data, relying on catastrophic forgetting without requiring logit access.
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
EEPO uses sample-then-forget rollouts with adaptive unlearning to boost exploration in RLVR, delivering relative gains of 24.3% on Qwen2.5-3B, 33.0% on Llama3.2-3B-Instruct, and 10.4% on Qwen3-8B-Base over GRPO across five reasoning benchmarks.
citing papers explorer
-
Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning
NPO enables stable unlearning of 50%+ training data in LLMs on TOFU by making collapse exponentially slower than gradient ascent, preserving sensible outputs where prior methods fail.
-
TRACER: Token ReAssignment for Concept ERasure in Generative Recommendation
TRACER uses token reassignment for concept-related items plus a coherence regularizer to unlearn specific concepts in generative recommendation while preserving utility better than baselines.
-
Model Unlearning Objectives Vary for Distinct Language Functions
Unlearning objectives should be tailored to distinct language functions, with a meta-learned RMU variant for dangerous knowledge and a multi-layer probe objective for toxicity, yielding strong results on four 7-8B models.
-
Representation-Guided Parameter-Efficient LLM Unlearning
REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.
-
Hey, That's My Data! Token-Only Dataset Inference in Large Language Models
CatShift detects training data membership in LLMs by comparing output shifts induced by fine-tuning on member versus non-member data, relying on catastrophic forgetting without requiring logit access.
-
Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
-
EEPO: Exploration-Enhanced Policy Optimization via Sample-Then-Forget
EEPO uses sample-then-forget rollouts with adaptive unlearning to boost exploration in RLVR, delivering relative gains of 24.3% on Qwen2.5-3B, 33.0% on Llama3.2-3B-Instruct, and 10.4% on Qwen3-8B-Base over GRPO across five reasoning benchmarks.
- RepSelect: Robust LLM Unlearning via Representation Selectivity