AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Return-to-Go Is More Than a Number: Q-Guided Alignment for Return-Conditioned Supervised Learning

Yuxiao Yang, Weitong Zhang

ICML 2026 · Reinforcement learning › RL from feedback · approach: RL fine-tuning of generative models · 0 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 Q-guided RTG-to-behavior alignment (Q-align DT vs QCS)Decision Transformer · D4RL Gym (sum of 9)normalized score: 812.3 -> 856.9Improved
add Q-guided RTG-to-behavior alignment (Q-align DT vs DT)Decision Transformer · D4RL Gym alignment (mean of 9)alignment RMSE: 21.39 -> 9.01Improved
add Q-guided alignment (Q-align DT vs QCS)Decision Transformer · D4RL AntMazenormalized score: antmaze-medium-diverse 79.5 -> 76.2; antmaze-medium-play 81.6 -> 85.6Mixed
add 1D convolution layer (ablation)Q-align DT · Gym Mediumalignment RMSE: 14.22 -> 9.19Improved
add 1D convolution layer (ablation)Q-align DT · Gym Medium-Replayalignment RMSE: 13.37 -> 11.31Improved
add 1D convolution layer (ablation)Q-align DT · Gym Medium-Expertalignment RMSE: 10.70 -> 6.80Improved
add 1D convolution layer (ablation)Q-align DT · Gym Medium-Replay/Medium/Medium-Expertnormalized score: 86.88 -> 86.45; 87.21 -> 87.70; 110.94 -> 111.37No clear effect
increase RTG perturbation offset Delta RTG from 0 to 10 (ablation)Q-align DT · halfcheetah-mediumD4RL scoreImproved
increase RTG perturbation offset Delta RTG from 0 to 10 (ablation)Q-align DT · halfcheetah-mediumalignment errorHurt
increase number of calibration samples from 5 to 40 (ablation)Q-align DT · HalfCheetah-vel zero-shotaverage episode return: -158.5 -> -142.3Improved
use Q-align DT in zero-shot transfer instead of QCSQ-align DT · HalfCheetah-vel zero-shotaverage episode return: -1298.7 -> -142.3Improved

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