THRD introduces a training-free multi-turn defense framework that models temporal risk accumulation to reduce jailbreak attack success rates to 0.2-4.0% on LLMs with under 1.5% utility degradation.
C ode A ttack: Revealing Safety Generalization Challenges of Large Language Models via Code Completion
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
An empirical study of 547 confirmed safety incidents from GitHub and literature derives a 33-type taxonomy showing constraint violations, destructive actions, and deception dominate in everyday coding-agent use.
SafeMoE isolates unsafe knowledge in domain-specific LoRA experts and routes them via a lightweight gate trained on safe responses to produce safer and more informative LLM outputs with zero-shot generalization.
citing papers explorer
-
THRD: A Training-Free Multi-Turn Defense Framework for Jailbreak Attacks on Large Language Models
THRD introduces a training-free multi-turn defense framework that models temporal risk accumulation to reduce jailbreak attack success rates to 0.2-4.0% on LLMs with under 1.5% utility degradation.
-
What Breaks When LLMs Code? Characterizing Operational Safety Failures of Agentic Code Assistants
An empirical study of 547 confirmed safety incidents from GitHub and literature derives a 33-type taxonomy showing constraint violations, destructive actions, and deception dominate in everyday coding-agent use.
-
Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
SafeMoE isolates unsafe knowledge in domain-specific LoRA experts and routes them via a lightweight gate trained on safe responses to produce safer and more informative LLM outputs with zero-shot generalization.