AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Self-Predictive Representations for Combinatorial Generalization in Behavioral Cloning

Daniel Lawson, Adriana Hugessen, Charlotte Cloutier, Glen Berseth, Khimya Khetarpal

ICLR 2026 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Zero-shot robustness and transfer · 9 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 BYOL-gamma auxiliary representation loss to GCBCGCBC · OGBench stitch average-allsuccess rate (%): 26 -> 35Improved
add minimal 1-step BYOL auxiliary lossGCBC · OGBench stitch average-allsuccess rate (%): 26 -> 26No clear effect
add BYOL-gamma on visual environmentsGCBC · OGBench average-visualsuccess rate (%): 34 -> 37Improved
add TD-SR auxiliary lossGCBC · OGBench average-visualsuccess rate (%): 34 -> 31Hurt
use predictive horizon gamma>0 instead of gamma=0 (ablation)BYOL-gamma · OGBench stitch average-allsuccess rate (%): 24 -> 33Improved
use predictive horizon gamma>0 instead of gamma=0 (ablation)BYOL-gamma · humanoidmaze-medium-stitchsuccess rate (%): 18 -> 54Improved
use cross-entropy loss instead of normalized squared l2 (ablation)BYOL-gamma · OGBench stitch average-allsuccess rate (%): 31 -> 33Improved
add action conditioning to the forward predictor (ablation)BYOL-gamma · OGBench stitch average-allsuccess rate (%): 33 -> 33No clear effect
add backward predictor (ablation)BYOL-gamma · OGBench stitch average-allsuccess rate (%): 33 -> 33No clear effect
use BYOL-gamma instead of TD-SR on large-state environmentsTD-SR · humanoidmaze-medium-stitchsuccess rate (%): 42 -> 51Improved
use TD-SR instead of BYOL-gamma on small antmazeBYOL-gamma · antmaze-large-stitchsuccess rate (%): 23 -> 19Hurt

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