AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Inherently Explainable Reinforcement Learning in Natural Language

XIANGYU PENG, Mark Riedl, Prithviraj Ammanabrolu

NeurIPS 2022 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Language-grounded reinforcement agents · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use HEX-RL hierarchical graph attention agent instead of baselinesHEX-RL (game and IM) · Jericho 9 games% completion (max score): 34.9 -> 35.8Improved
add intrinsic motivation reward to HEX-RL (ablation)HEX-RL · Jericho 9 games% completion (final score): 27.2 -> 28.2Improved
compare HEX-RL game and IM against Q*BERT on final scoreHEX-RL · Jericho 9 games% completion (final score): 30.8 -> 28.2Hurt
use Hierarchical Graph Attention explanations instead of LSTM attentionHEX-RL · zork1, library, balanceshuman preference (confidence, understandability)Improved
use Hierarchical Graph Attention explanations instead of LSTM attentionHEX-RL · zork1, library, balanceshuman-likeness preferenceMixed
use temporal instead of immediate explanationsHEX-RL · human judgmentgoal contextImproved
use temporal instead of immediate explanationsHEX-RL · human judgmentcoherence and readabilityNo clear effect
add CALM language model filter to Bayes filter (ablation)temporal explanations · human evaluationconfidence and understandabilityImproved
add semantic filter to Bayes+CALM (ablation)temporal explanations · human evaluationconfidence, human-likeness, understandabilityImproved
use HEX-RL with intrinsic motivation instead of Q*BERTHEX-RL · pentariaverage score: 50 -> 44.7Hurt
use HEX-RL with intrinsic motivation instead of Q*BERTHEX-RL · detectiveaverage score: 274 -> 276.9No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy