3 papers across 2 sessions
We offer statistically robust methods for preference learning that leverage response time in the estimation of rewards to yield large improvements in statistical efficiency.