AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Learning Tree Interpretation from Object Representation for Deep Reinforcement Learning

Guiliang Liu, Xiangyu Sun, Oliver Schulte, Pascal Poupart

NeurIPS 2021 · Probabilistic and scientific modeling › Molecules and proteins · approach: Probabilistic embeddings and relational inference · 7 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 IMONet latent features instead of VAE latent features (ablation)MCRTS · Flappy BirdVR: 7.83E-2 -> 8.53E-2Improved
use IMONet latent features instead of VAE latent features (ablation)MCRTS · Space InvadersVR: 4.82E-2 -> 5.37E-2Improved
use IMONet latent features instead of VAE latent features (ablation)MCRTS · AssaultVR: 6.58E-2 -> 7.53E-2Improved
replace CART with MCRTS split search on IMONet features (ablation)IMONet tree · Flappy BirdVR-PL: 4.02E-4 -> 1.37E-3Improved
replace CART with MCRTS split search on IMONet features (ablation)IMONet tree · AssaultNumber of leaves: 315 -> 83Improved
use IMONet features with MCRTS instead of raw-input CART (ablation)tree mimic · Flappy BirdNumber of leaves: 1007 -> 62Improved
use IMONet features with MCRTS instead of raw-input CART (ablation)tree mimic · Space InvadersVariance Reduction (VR): 4.96E-2 -> 5.37E-2Mixed
use IMONet+MCRTS instead of raw-input M5-RT (ablation)tree mimic · Flappy BirdVariance Reduction (VR): 9.59E-2 -> 8.53E-2Hurt
use VAE features instead of raw input for CART (ablation)CART · Flappy BirdVariance Reduction (VR): 8.51E-2 -> 7.25E-2Hurt
use IMONet features instead of VAE for VR-LMT (ablation)VR-LMT · Space InvadersVariance Reduction (VR): 3.52E-2 -> 4.54E-2Improved
MCRTS with look-ahead split vs greedy linear model tree (ablation)IMONet tree · AssaultVR-PL: 2.27E-4 -> 9.07E-4Improved
use latent mimic tree instead of super-pixel tree for human interpretabilityIMONet+MCRTS · Human evaluationpreference (%): 83Improved

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