Machine unlearning conflates reversing the influence of specific training examples (untraining) with removing the full underlying distribution or behavior (unlearning).
Open problems in machine unlearning for ai safety
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
LLM unlearning is reframed as inadvertently installing backdoor triggers on forget-tokens; Random Noise Augmentation is introduced as a defense that improves robustness with theoretical guarantees.
UVR is a training-free framework that uses attention modulation based on identified information flow stages in multimodal DiT attention to erase unsafe semantics in image synthesis and editing at 91% and 77% rates while preserving quality.
Standard unlearning raises false-negative rates and clinical risk in medical image models, but a modified SalUn-CRA variant using entropy-based forgetting for malignant samples matches full retraining risk while still removing data influence.
Parameter-difference and model-inversion attacks can identify forgotten classes after machine unlearning on standard image datasets.
Machine unlearning in LLMs is often reversible via fine-tuning, indicating suppression not deletion, and a new representation-level framework identifies four forgetting regimes based on reversibility and catastrophicity.
Physical Foundation Models are fixed physical hardware realizations of foundation-scale neural networks that compute via inherent material dynamics, potentially delivering orders-of-magnitude gains in energy efficiency, speed, and density over digital systems.
MoralityGym is a new benchmark using 98 ethical dilemmas in sequential environments to evaluate hierarchical moral alignment in AI agents via Morality Chains and a Morality Metric.
citing papers explorer
-
Is your algorithm unlearning or untraining?
Machine unlearning conflates reversing the influence of specific training examples (untraining) with removing the full underlying distribution or behavior (unlearning).
-
Improving LLM Unlearning Robustness via Random Perturbations
LLM unlearning is reframed as inadvertently installing backdoor triggers on forget-tokens; Random Noise Augmentation is introduced as a defense that improves robustness with theoretical guarantees.
-
Unified Safe In-context Image Generation in Multimodal Diffusion Transformers via Restricting Unsafe Information Flows
UVR is a training-free framework that uses attention modulation based on identified information flow stages in multimodal DiT attention to erase unsafe semantics in image synthesis and editing at 91% and 77% rates while preserving quality.
-
Does Machine Unlearning Preserve Clinical Safety? A Risk Analysis for Medical Image Classification
Standard unlearning raises false-negative rates and clinical risk in medical image models, but a modified SalUn-CRA variant using entropy-based forgetting for malignant samples matches full retraining risk while still removing data influence.
-
Label Leakage Attacks in Machine Unlearning: A Parameter and Inversion-Based Approach
Parameter-difference and model-inversion attacks can identify forgotten classes after machine unlearning on standard image datasets.
-
Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs
Machine unlearning in LLMs is often reversible via fine-tuning, indicating suppression not deletion, and a new representation-level framework identifies four forgetting regimes based on reversibility and catastrophicity.
-
Physical Foundation Models: Fixed hardware implementations of large-scale neural networks
Physical Foundation Models are fixed physical hardware realizations of foundation-scale neural networks that compute via inherent material dynamics, potentially delivering orders-of-magnitude gains in energy efficiency, speed, and density over digital systems.
-
MoralityGym: A Benchmark for Evaluating Hierarchical Moral Alignment in Sequential Decision-Making Agents
MoralityGym is a new benchmark using 98 ethical dilemmas in sequential environments to evaluate hierarchical moral alignment in AI agents via Morality Chains and a Morality Metric.