AI Evaluation Awareness
Matthew Berman discusses how AI models can recognize when they are being evaluated and the risks associated with this awareness.
Read More
Research
Reinforcement Pre-Training in AI
Matthew Berman discusses Reinforcement Pre-Training (RPT), a new approach to AI that enhances learning capabilities without curated datasets.
Read More
Claim Extraction from LLMs
Dasha Metropolitansky presents Claimify, a new tool for extracting verifiable claims from language model outputs, enhancing fact-checking processes.
Read More