RACC defines six representation-aware coverage criteria that score jailbreak test suites by measuring activation of safety concepts extracted from LLM hidden states on a calibration set.
Advanc- ing llm safe alignment with safety representation ranking
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 2
method 1
citation-polarity summary
verdicts
UNVERDICTED 6representative citing papers
The work introduces a distributional view of visual mechanistic interpretability that casts the task as KL-minimal optimization and realizes it through a soft-constraint principle implemented with energy-guided diffusion posterior sampling on models such as DINOv3.
MORA breaks the safety-helpfulness ceiling in LLMs by pre-sampling single-reward prompts and rewriting them to incorporate multi-dimensional intents, delivering 5-12.4% gains in sequential alignment and 4.6% overall improvement in simultaneous alignment.
DMI-Lib delivers 0.4-6.8% overhead for offline batch LLM inference and ~6% for moderate online serving while exposing rich internal signals across backends, cutting latency overhead 2-15x versus prior observability baselines.
Precise Shield identifies safety neurons in VLLMs via activation contrasts and aligns only them with gradient masking, boosting safety, preserving generalization, and enabling zero-shot cross-lingual and cross-modal transfer.
ReGA uses safety-critical representations to guide abstraction in model-based analysis, enabling scalable detection of harmful LLM inputs with reported AUROC of 0.975 at prompt level.
citing papers explorer
-
RACC: Representation-Aware Coverage Criteria for LLM Safety Testing
RACC defines six representation-aware coverage criteria that score jailbreak test suites by measuring activation of safety concepts extracted from LLM hidden states on a calibration set.
-
A Distributional View for Visual Mechanistic Interpretability: KL-Minimal Soft-Constraint Principle
The work introduces a distributional view of visual mechanistic interpretability that casts the task as KL-minimal optimization and realizes it through a soft-constraint principle implemented with energy-guided diffusion posterior sampling on models such as DINOv3.
-
Explaining and Breaking the Safety-Helpfulness Ceiling via Preference Dimensional Expansion
MORA breaks the safety-helpfulness ceiling in LLMs by pre-sampling single-reward prompts and rewriting them to incorporate multi-dimensional intents, delivering 5-12.4% gains in sequential alignment and 4.6% overall improvement in simultaneous alignment.
-
Enabling Performant and Flexible Model-Internal Observability for LLM Inference
DMI-Lib delivers 0.4-6.8% overhead for offline batch LLM inference and ~6% for moderate online serving while exposing rich internal signals across backends, cutting latency overhead 2-15x versus prior observability baselines.
-
Precise Shield: Explaining and Aligning VLLM Safety via Neuron-Level Guidance
Precise Shield identifies safety neurons in VLLMs via activation contrasts and aligns only them with gradient masking, boosting safety, preserving generalization, and enabling zero-shot cross-lingual and cross-modal transfer.
-
ReGA: Model-Based Safeguard for LLMs via Representation-Guided Abstraction
ReGA uses safety-critical representations to guide abstraction in model-based analysis, enabling scalable detection of harmful LLM inputs with reported AUROC of 0.975 at prompt level.