Current: Home Research

Research

Contrasting two Bayesian revision theories

It develops a preference revision theory in the Bayesian setting in an AGM-like style, called Bayesian preference revision theory. It then contrasts Bayesian preference revision theory with the classic Bayesian belief revision theory and presents both negative and postive results.

Strictly proper scoring rules for imprecise credences

It illutrates the meaning of the impossibility theorem purported in Seindenfeld et al., (2012) and offers a characterization of two diffrent kinds of strictly proper lexicographic-valued scoring rules for imprecise credences.

Accuracy dominance argument for Probabilism (with David McCarthy)

Let's keep the secret for a while :).

On incomplete preference for aligning LLMs using RLHF

Let's keep the secret for a while :). 

Bayes meets AGM (with David McCarthy)

Let's keep the secret for a while :).

Growing awareness and conservatism rule of updating

Let's keep the secret for a while :).



TOP