Fine-tuning updates frequently stale activation monitors for language model safety while quantization does not, with degradation predictable and repairable via label-free realignment.
Alignguard-lora: Alignment- presving fine-tuning via fisher-guided decomposition and riemannian-geodesic collision regularization.arXiv preprint arXiv:2508.02079
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
RefusalGuard constrains updates in hidden representation space to preserve safety-relevant geometric structure during fine-tuning, maintaining low attack success rates on safety benchmarks while preserving task performance.
citing papers explorer
-
Do Activation Monitors Survive Model Updates? Benchmarking, Predicting, and Repairing Activation-Monitor Staleness
Fine-tuning updates frequently stale activation monitors for language model safety while quantization does not, with degradation predictable and repairable via label-free realignment.
-
RefusalGuard: Geometry-Preserving Fine-Tuning for Safety in LLMs
RefusalGuard constrains updates in hidden representation space to preserve safety-relevant geometric structure during fine-tuning, maintaining low attack success rates on safety benchmarks while preserving task performance.