LLM Safety From Within: Detecting Harmful Content with Internal Representations
Paper • 2604.18519 • Published • 26
Haolun Wu
haolun-wu
AI & ML interests
Knowledge modeling, storage, and retrieval.
Recent Activity
authored a paper 20 days ago
Can LLM Agents Really Debate? A Controlled Study of Multi-Agent Debate in Logical Reasoning authored a paper 20 days ago
PartnerMAS: An LLM Hierarchical Multi-Agent Framework for Business Partner Selection on High-Dimensional Features authored a paper 20 days ago
Logits are All We Need to Adapt Closed ModelsOrganizations
