AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

Difference-Aware Retrieval Polices for Imitation Learning

Quinn Pfeifer, Ethan Pronovost, Paarth Shah, Khimya Khetarpal, Siddhartha Srinivasa, Abhishek Gupta

ICLR 2026 · Reinforcement learning › Imitation learning · approach: Offline imitation from imperfect data

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 difference-aware retrieval policy instead of plain behavior cloningBC (MLP) · Robosuite and RoboCasa (low-dim state)success rate (%): Stack 47 -> 72Improved
use DARP instead of BC on vision-based tasks (R3M features)BC · Robosuite Pegsuccess rate (%): 17 -> 52Improved
add neighbor difference vectors (query vs neighbor state) to the predictor input (ablation)DARP · Robosuite ablationsuccess rateImproved
use permutation-invariant aggregation of neighbor predictions (ablation)DARP · Robosuite ablationsuccess rateImproved
include neighbor actions in predictor input (ablation)DARP · Robosuite ablationsuccess rateImproved
select neighbors by distance rather than at random (ablation)DARP · Robosuite ablationsuccess rateImproved
implicit neighbor aggregation in the architecture (DARP) instead of explicit smoothness loss (MRIL) (ablation)MRIL · MuJoCo HalfCheetahreturn: 701.58 -> 5515.41Improved
use pure non-parametric retrieval (R&P, LWR)BC · MuJoCoreturn: Hopper 2313.65 -> 711.82 (R&P)Hurt
replace Euclidean distance with cosine similarity for neighbor retrieval (ablation)DARP · Robosuite Stack (R3M)success rate (%): 75 -> 70No clear effect
combine DARP with a diffusion policy backbone (ablation)DARP (MLP) · Robosuite Stacksuccess rate (%): 62 -> 82Improved
use DARP with GMM head on multimodal Push-TBC · Push-Tperformance: 22% improvement over BCImproved
use DARP on discontinuous modified D4RL UmazeBC · D4RL Umaze (modified)success rate: 25% -> 57%Improved

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