arXiv Machine Learning
Decoy Direction Optimization: A Post-Hoc Defense Against LLM Abliteration
Read the original on arXiv Machine Learning →The Flow has not summarised this story yet — read it at arXiv Machine Learning.
Related stories
arXiv AI
Aug 17
Tripwire: Triggering Aligned Refusal via Statistically Certified Safety Neurons
arXiv:2608. 14392v1 Announce Type: new Abstract: Neuron- and path-level interventions offer the finest-grained route to defending large language models (LLMs) against jailbreak attacks, yet existing methods fall short of this promise, i.
arXiv AI
Sep 10
Bait-and-Recover: Poisoning Internal Refusal Signals to Defend LLMs against White-Box Editing Jailbreaks
arXiv:2609.05794v1 Announce Type: cross Abstract: Open-weight large language models face a low-cost white-box threat from representation engineering attacks. Attackers can estimate refusal directions...
arXiv AI
Jul 20
Latent Fusion Jailbreak: Blending Harmful and Harmless Representations to Elicit Unsafe LLM Outputs
arXiv:2508. 10029v3 Announce Type: replace-cross Abstract: Safety-aligned large language models can still be manipulated through white-box interventions that modify their internal representations.
Hugging Face Trending Papers
Jul 22
Defense Against LLM Backdoors using Critical Neuron Isolation Pruning
Large language models (LLMs) are vulnerable to backdoor attacks, where hidden triggers induce malicious outputs. Existing defenses generally fall into inference-time detection or training-time mitigation, but face two key limitations.
arXiv AI
Jul 9
NonTextual Target Attack
arXiv:2510. 02999v5 Announce Type: replace-cross Abstract: Existing gradient-based jailbreak attacks on Large Language Models (LLMs) typically optimize adversarial suffixes to align the LLM output with predefined target responses.
arXiv AI
Jul 23
Defense Against LLM Backdoors using Critical Neuron Isolation Pruning
arXiv:2607. 19894v1 Announce Type: cross Abstract: Large language models (LLMs) are vulnerable to backdoor attacks, where hidden triggers induce malicious outputs.