MetaSC: Test-Time Safety Specification Optimization for Language Models
We propose a novel dynamic safety framework that optimizes language model (LM) safety reasoning at inference time without modifying model weights. Building on recent advances in self-critique methods, our approach leverages a meta-critique mechanism that iteratively updates safety prompts-termed specifications-to drive the critique and revision process adaptively. This test-time optimization not only improves performance against adversarial jailbreak requests but also in diverse general safety-related tasks, such as avoiding moral harm or pursuing honest responses. Our empirical evaluations across several language models demonstrate that dynamically optimized safety prompts yield significantly higher safety scores compared to fixed system prompts and static self-critique defenses. Code to be released at https://github.com/vicgalle/meta-self-critique.git .
Code (1)
Tasks
Language ModelingLanguage ModellingSimilar Papers 제목 키워드 기반
MetaScale: Test-Time Scaling with Evolving Meta-Thoughts
One critical challenge for large language models (LLMs) for making complex reasoning is their reliance on matching reasoning patterns from training data, instead of proactively selecting the most appropriate cognitive st…
Verifying Controllers Against Adversarial Examples with Bayesian Optimization
Recent successes in reinforcement learning have lead to the development of complex controllers for real-world robots. As these robots are deployed in safety-critical applications and interact with humans, it becomes crit…
Bayesian Optimizationreinforcement-learningReinforcement LearningReinforcement Learning (RL)Counterexample Guided RL Policy Refinement Using Bayesian Optimization
Constructing Reinforcement Learning (RL) policies that adhere to safety requirements is an emerging field of study. RL agents learn via trial and error with an objective to optimize a reward signal. Often policies that a…
Bayesian OptimizationReinforcement Learning (RL)Reasoning over Boundaries: Enhancing Specification Alignment via Test-time Deliberation
Large language models (LLMs) are increasingly applied in diverse real-world scenarios, each governed by bespoke behavioral and safety specifications (spec) custom-tailored by users or organizations. These spec, categoriz…
Adaptive Approach to Enhance Machine Learning Scheduling Algorithms During Runtime Using Reinforcement Learning in Metascheduling Applications
Metascheduling in time-triggered architectures has been crucial in adapting to dynamic and unpredictable environments, ensuring the reliability and efficiency of task execution. However, traditional approaches face signi…
Reinforcement Learning