Reasoning Models Don't Always Say What They Think
Yanda Chen, Joe Benton, Ansh Radhakrishnan, Jonathan Uesato, Carson Denison, John Schulman, Arushi Somani, Peter Hase and 7 more
Who it is from
Anthropicaffiliation declared by the authors, according to OpenAlex
Other papers from the same organizations
- Collective Constitutional AI: Aligning a Language Model with Public InputAnthropic · 257 citations
- SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model InterpretabilityAnthropic · 194 citations
- Language models transmit behavioural traits through hidden signals in dataAnthropic · 193 citations
- Phantom: General Backdoor Attacks on Retrieval Augmented Language GenerationAnthropic · 156 citations
- The Assistant Axis: Situating and Stabilizing the Default Persona of Language ModelsAnthropic · 149 citations
- SoK: Watermarking for AI-Generated ContentAnthropic · 134 citations