Research
Find relevant papers written by the Pareto team
A community-driven vision for a new knowledge resource for AI
Synthesising insights from over 50 researchers at an AAAI workshop to outline a new open engineering framework for knowledge modules in AI.
Debating with More Persuasive LLMs Leads to More Truthful Answers
Evaluating debate as a scalable method for aligning AI models, showing that structured argumentation between LLMs helps humans and weaker models identify truth more reliably.