🏢 Department of Statistics, University of Chicago
Conformal Alignment: Knowing When to Trust Foundation Models with Guarantees
·3577 words·17 mins·
loading
·
loading
Natural Language Processing
Question Answering
🏢 Department of Statistics, University of Chicago
Conformal Alignment certifies trustworthy foundation model outputs by guaranteeing a user-specified fraction meet alignment criteria, regardless of the model or data.
Building a stable classifier with the inflated argmax
·2014 words·10 mins·
loading
·
loading
AI Generated
AI Theory
Fairness
🏢 Department of Statistics, University of Chicago
Boost classifier stability with the novel inflated argmax, guaranteeing reliable multiclass classification without distributional assumptions!
BoNBoN Alignment for Large Language Models and the Sweetness of Best-of-n Sampling
·2612 words·13 mins·
loading
·
loading
AI Generated
Natural Language Processing
Large Language Models
🏢 Department of Statistics, University of Chicago
BoNBON alignment optimizes large language model (LLM) outputs towards human preferences using best-of-n sampling, maximizing win-rate against base models with minimal off-target impact.