Author

W. B. Knox

h-index 24 View on Semantic Scholar

We have 2 of 51 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Aug 2026

A Framework for Designing Reward Functions: From Objectives to Features to Human-Aligned Reward Functions

To the best of the knowledge, this is the first reward-design method that maintains a deterministically conflict-free feasible weight region, narrowed to a desired tolerance via a separation oracle with O(n log \kappa) preference queries.

D. Shi, W. B. Knox · 0 citations

#artificial intelligence Preprint Sep 2026

Specifying Reward Functions for RL Without Environment Sampling

This work introduces Experience-Free Autonomous Reward Specification (EARS), a method for learning reward functions from preferences without environment interaction and compares EARS to baselines that also enable reward specification without environment interaction--namely, methods that directly prompt an LLM to genera...

Stephane Hatgis-Kessell, W. B. Knox, Emma Brunskill · 1 citation