REP elicits hidden LLM reasoning traces via in-context shadow demonstrations, raising similarity to internal traces while retaining distillation utility across datasets and models.
Salt: Steering activations towards leakage-free thinking in chain of thought
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
No existing AI security framework covers a majority of the 193 identified multi-agent system threats in any category, with OWASP Agentic Security Initiative achieving the highest overall coverage at 65.3%.
Steering Llama-2-7B-Chat and Qwen2.5-7B-Instruct teachers and distilling students on benign data transfers measurable jailbreak susceptibility, with Llama showing threshold behavior at α = -0.15 and Qwen reaching transfer ratios up to 0.61.
citing papers explorer
-
Security Considerations for Multi-agent Systems
No existing AI security framework covers a majority of the 193 identified multi-agent system threats in any category, with OWASP Agentic Security Initiative achieving the highest overall coverage at 65.3%.