English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Large Language Models (LLMs) are now widely used in various applications,making it crucial to align their ethical standards with human values. However,recent jail-breaking methods demonstrate that this alignment can be underminedusing carefully constructed prompts. In our study, we reveal a new threat toLLM alignment when a bad actor has access to the model's output logits, acommon feature in both open-source LLMs and many commercial LLM APIs (e.g.,certain GPT models). It does not rely on crafting specific prompts. Instead, itexploits the fact that even when an LLM rejects a toxic request, a harmfulresponse often hides deep in the output logits. By forcefully selectinglower-ranked output tokens during the auto-regressive generation process at afew critical output positions, we can compel the model to reveal these hiddenresponses. We term this process model interrogation. This approach differs fromand outperforms jail-breaking methods, achieving 92% effectiveness compared to62%, and is 10 to 20 times faster. The harmful content uncovered through ourmethod is more relevant, complete, and clear. Additionally, it can complementjail-breaking strategies, with which results in further boosting attackperformance. Our findings indicate that interrogation can extract toxicknowledge even from models specifically designed for coding tasks.

Make Them Spill the Beans! Coercive Knowledge Extraction from (Production) LLMs