Measuring the persuasiveness of language models
ID: 006b10e3-7665-5361-af0e-749afa0511ff
STIX ID: report--006b10e3-7665-5361-af0e-749afa0511ff
Feed Name: Anthropic Research
This Anthropic research post reports an experimental study measuring how persuasive various Claude language models are compared to human writers across 28 topics and 56 claims, finding a scaling trend with model capability and that Claude 3 Opus approaches human-level persuasiveness; it details the experimental design, prompting strategies, results, limitations, ethical considerations, and releases the underlying dataset.
Your team is not currently subscribed to this feed. You must subscribe to it in order to see this post.