Contributed Oral Talks
Abstract
• Emergent Misalignment as a Distributional Phase Transition in Rank-1 LoRA Fine-Tuning • Where did the ambiguity go? Examining how multimodal models interpret polysemous words • How Much Do Circuits Tell Us? Measuring the Consistency and Specificity of Language Model Circuits • The Future of Facts: Tracing the Factual Generation-Verification Gap
Chat is not available.
Successful Page Load