AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Contrastive Reinforcement Learning of Symbolic Reasoning Domains

Gabriel Poesia, WenXin Dong, Noah Goodman

NeurIPS 2021 · Reinforcement learning › Visual and representation RL · approach: Inferring decision-making from behavior · 4 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 InfoNCE contrastive loss instead of value regression (CVI) (ablation)ConPoLe · CommonCore Equationssuccess rate (%): 89.0% -> 92.5%Improved
use InfoNCE contrastive loss instead of value regression (CVI) (ablation)ConPoLe · CommonCore Sortingsuccess rate (%): 77.0% -> 100.0%Improved
use InfoNCE contrastive loss instead of value regression (CVI) (ablation)ConPoLe · CommonCore Fractionssuccess rate (%): 84.5% -> 96.0%Improved
use InfoNCE contrastive loss instead of value regression (CVI) (ablation)ConPoLe · CommonCore Multiplicationsuccess rate (%): 100.0% -> 99.5%No clear effect
use only local negatives (ConPoLe-local) instead of all beam-search negatives (ablation)ConPoLe · CommonCore Equationssuccess rate (%): 92.5% -> 76.5%Hurt
use only local negatives (ConPoLe-local) instead of all beam-search negatives (ablation)ConPoLe · CommonCore Fractionssuccess rate (%): 96.0% -> 86.0%Hurt
use only local negatives instead of all negatives (ablation)ConPoLe · CommonCore Sortingsuccess rate (%): 100.0% -> 100.0%No clear effect
use only local negatives instead of all negatives (ablation)CVI · CommonCoresuccess rateNo clear effect
replace 1-step lookahead with bounded BFS (ablation)ADI and DAVI · CommonCoresuccess rateNo clear effect
train with contrastive examples (ConPoLe) instead of behavioral cloningConPoLe · CommonCore Equationssuccess rate (%): 4.5% -> 92.5%Improved
train with contrastive examples (ConPoLe) instead of DRRNConPoLe · CommonCore Equationssuccess rate (%): 2.5% -> 92.5%Improved
use ConPoLe representations instead of edit distance1-NN classifier · Khan Academy equation categoriesaccuracy: 0.714 -> 0.905Improved

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