Author

Emma Beharry

h-index 1 View on Semantic Scholar

We have 1 of 2 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.

Review

GoodLiars: A Multi-Turn Extension of Reinforcement Learning-Based Belief Disruption

This work extends GOODLIAR from a single-turn attack to a multi-turn one and test two algorithms: a multi-turn variant of GOODLIAR’s DA-ILQL, an offline RL method with on-policy data aggregation, and a Proximal Policy Optimization (PPO) attacker, which is adopted because its per-turn rewards better match a multi-turn conversation.

Emma Beharry, Elizabeth Gallagher, Abel John · 0 citations