Toxic context can be laundered into memory summaries that stay below toxicity thresholds while still driving higher downstream toxicity in LLM agents compared to neutral baselines.
Ai safety in generative ai large language models: A survey
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 4roles
background 1polarities
background 1representative citing papers
BarrierSteer applies control barrier functions to LLM latent states for constraint-guided steering that reduces unsafe generations while preserving utility.
An accurate and trusted AI system cannot achieve human-level reasoning because there exist tasks easily solvable by humans but not by the system.
Survey of harmful fine-tuning attacks on LLMs, their variants, defense strategies, mechanical analysis, and evaluation methodologies.
citing papers explorer
-
State Contamination in Memory-Augmented LLM Agents
Toxic context can be laundered into memory summaries that stay below toxicity thresholds while still driving higher downstream toxicity in LLM agents compared to neutral baselines.
-
BarrierSteer: LLM Safety via Learning Barrier Steering
BarrierSteer applies control barrier functions to LLM latent states for constraint-guided steering that reduces unsafe generations while preserving utility.
-
Limitations on Accurate, Trusted, Human-level Reasoning
An accurate and trusted AI system cannot achieve human-level reasoning because there exist tasks easily solvable by humans but not by the system.
-
Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey
Survey of harmful fine-tuning attacks on LLMs, their variants, defense strategies, mechanical analysis, and evaluation methodologies.