Researchers Pursue New Ways to Make AI More Truthful, Reasonable, and Human-Aware
-
Jacob's research uses game theory to improve natural language model output and make AI systems more reliable and truthful.
-
Shen's project aims to calibrate language models when they are poorly calibrated, tuning the confidence output.
-
Butoi's team is creating techniques to allow vision-language models to better reason about what they're seeing and understand key phrases.
-
Peng's group is developing embodied AI models to assist people with physical constraints in a simulated environment.
-
Peng wants to build systems that convey human knowledge and behave in understandable, human-like ways.