REVIEW 2 cited by
Stealthy and Persistent Unalignment on Large Language Models via Backdoor Injections
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Recent developments in Large Language Models (LLMs) have manifested significant advancements. To facilitate safeguards against malicious exploitation, a body of research has concentrated on aligning LLMs with human preferences and inhibiting their generation of inappropriate content. Unfortunately, such alignments are often vulnerable: fine-tuning with a minimal amount of harmful data can easily unalign the target LLM. While being effective, such fine-tuning-based unalignment approaches also have their own limitations: (1) non-stealthiness, after fine-tuning, safety audits or red-teaming can easily expose the potential weaknesses of the unaligned models, thereby precluding their release/use. (2) non-persistence, the unaligned LLMs can be easily repaired through re-alignment, i.e., fine-tuning again with aligned data points. In this work, we show that it is possible to conduct stealthy and persistent unalignment on large language models via backdoor injections. We also provide a novel understanding on the relationship between the backdoor persistence and the activation pattern and further provide guidelines for potential trigger design. Through extensive experiments, we demonstrate that our proposed stealthy and persistent unalignment can successfully pass the safety evaluation while maintaining strong persistence against re-alignment defense.
Forward citations
Cited by 2 Pith papers
-
Paladin: Defending LLM-enabled Phishing Emails with a New Trigger-Tag Paradigm
A trigger-tag watermark embedded by fine-tuning lets modified LLMs mark their own phishing outputs for cheap detection.
-
Mechanistic Exploration of Backdoored Large Language Model Attention Patterns
Backdoored Qwen2.5-3B models show attention deviations concentrated in layers 20-30, with single-token triggers localized and multi-token triggers diffuse.
Discussion (0). Continue with ORCID to comment.