AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

The LoCA Regret: A Consistent Metric to Evaluate Model-Based Behavior in Reinforcement Learning

Harm Van Seijen, Hadi Nekoei, Evan Racah, Sarath Chandar

NeurIPS 2020 · Reinforcement learning › Visual and representation RL · approach: Latent-space world models · 5 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
add LoCA pretrainingMB-VI · Tabular gridworldregret (x1000): 7.29 -> 0.00Improved
add LoCA pretrainingSarsa(0.95) · Tabular gridworldregret (x1000): 30.09 -> 14.92Improved
add LoCA pretrainingQ-learning · Tabular gridworldregret (x1000): 68.25 -> 39.52Improved
use value iteration instead of single value update for planning (ablation)MB-VI vs MB-SU · Tabular gridworldLoCA regret (x1000): 1.72 -> 0.00Improved
step-size multiplier 0.1 (poor hyperparameters) (ablation)MB-VI · Tabular gridworldLoCA regret (x1000): 0.00 -> 0.00No clear effect
step-size multiplier 0.1 (poor hyperparameters) (ablation)Sarsa(0.95) · Tabular gridworldLoCA regret (x1000): 14.92 -> 134.82Hurt
expand state space by 5x with irrelevant features (ablation)MB-VI · Tabular gridworldLoCA regret (x1000): 0.00 -> 0.00No clear effect
expand state space by 2x with irrelevant features (ablation)Sarsa(0.95) · Tabular gridworldLoCA regret (x1000): 14.92 -> 26.30Hurt
expand state space by 10x with irrelevant features (ablation)MB-SU · Tabular gridworldLoCA regret (x1000): 1.72 -> 15.67Hurt
use 2-step on-policy model instead of 1-step (ablation)tabular model-based method · Tabular gridworldLoCA regret (x1000): 1.68 -> 24.87Hurt
use 5-step on-policy model instead of 1-step (ablation)tabular model-based method · Tabular gridworldLoCA regret (x1000): 1.68 -> 34.77Hurt
use 5-step on-policy model instead of 1-step (ablation)tabular model-based method · Tabular gridworlddefault regret (x1000): 42.98 -> 34.51Improved

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