Faithful Dual-constrained Erasure for Robust LLM Safety Alignment

arXiv:2609.39279 · cs.CR, cs.AI · Submitted 2026-09-30 · Read on arXiv

cs.CR, cs.AI

Submitted: 2026-09-30

Updated: 2026-09-30

Terminology

• Alignment — 23× in this paper · explained in 1 episode
• Accuracy — 13× in this paper · explained in 1 episode
• Robustness — 13× in this paper · explained in 5 episodes
• Fine-Tuning — 12× in this paper · explained in 5 episodes
• Gradient Ascent — 10× in this paper
• Machine Unlearning — 10× in this paper · explained in 2 episodes
• Fisher Information — 6× in this paper · explained in 3 episodes
• Catastrophic Forgetting — 5× in this paper · explained in 5 episodes
• Constrained Knowledge Unlearning — 5× in this paper
• Benchmark — 4× in this paper · explained in 2 episodes
• Perplexity — 4× in this paper · explained in 3 episodes
• Regularization — 4× in this paper · explained in 3 episodes
• Ablation — 3× in this paper · explained in 1 episode
• Ablation study — 3× in this paper · explained in 1 episode
• Collapse of Irrelevant Representations — 3× in this paper
• Large Language Models (LLMs) — 3× in this paper · explained in 6 episodes
• Principle of Minimal Functional Intervention — 3× in this paper
• Adam — 2× in this paper · explained in 2 episodes
• Benchmarking — 2× in this paper · explained in 3 episodes
• Core — 2× in this paper · explained in 2 episodes
• Decoupling — 2× in this paper · explained in 2 episodes
• Fisher Information Matrix — 2× in this paper
• LLM-as-a-Judge — 2× in this paper · explained in 3 episodes
• Quantization — 2× in this paper · explained in 12 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Adversarial Attacks — 1× in this paper · explained in 1 episode
• Consistency — 1× in this paper · explained in 3 episodes
• Erasure of Language Memory — 1× in this paper
• Failure Modes — 1× in this paper · explained in 3 episodes
• Influence Functions — 1× in this paper · explained in 2 episodes
• Loss Function — 1× in this paper · explained in 2 episodes
• Parameter-Efficient Fine-Tuning — 1× in this paper · explained in 2 episodes
• Pruning — 1× in this paper · explained in 4 episodes
• Transformer — 1× in this paper · explained in 2 episodes

Sources

• Knowledge Unlearning for Mitigating Privacy Risks in Language Models
• Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!
• MUSE: Machine Unlearning Six-Way Evaluation for Language Models
• Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs
• From Cognition to Computation: A Comparative Review of Human Attention and Transformer Architectures

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs