AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Scheduling Thoughts: Learning the Order of Thought in Diffusion Language Models

Jiawei Xu, Minghui Liu, Aakriti Agrawal, Yifan Chen, Furong Huang

ICML 2026 · Large language models › Language, safety and interpretability · approach: Parallel and early-exit decoding · 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
use learned order policy instead of confidence orderingSMDM-1B Sudoku decoding · Sudoku testpuzzle accuracy (%): 82.0 -> 91.8Improved
use learned order policy instead of human expert orderSMDM-1B Sudoku decoding · Sudoku testpuzzle accuracy (%): 76.6 -> 91.8Improved
use confidence ordering instead of randomSMDM-1B Sudoku decoding · Sudoku testpuzzle accuracy (%): 62.7 -> 82.0Improved
second-stage fine-tuning of diffusion head along learned order (ablation)SAS order policy · Sudoku testpuzzle accuracy (%): 91.8 -> 97.5Improved
second-stage fine-tuning along human expert order (ablation)Human order · Sudoku testpuzzle accuracy (%): 76.6 -> 97.8Improved
replace default LLaDA schedule with learned order policyLLaDA-8B-Instruct · GSM8Kpass@1 (%): 64% -> 76%Improved
replace default schedule with learned order policyLLaDA-8B-Instruct · MBPPpass@1 (%): 39.5% -> 41%Improved
transfer order policy trained on code to math, full diffusion 256-256SAS policy · GSM8Kaccuracy: +8.5%Improved
transfer order policy trained on math to code, 256-256SAS policy · MBPPaccuracy: +1.5%Improved
transfer order policy trained on math to code, 256-32SAS policy · MBPPaccuracy: +0.33%Improved
dense self-aware reward instead of 0/1 terminal reward for order policy (ablation)SAS order policy · GSM8K (5-shot)accuracyMixed

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