AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Planned Diffusion

Daniel Mingyi Israel, Tian Jin, Ellie Y Cheng, Guy Van den Broeck, Aditya Grover, Suvinay Subramanian, Michael Carbin

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Masked diffusion language model decoding · 5 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 planned diffusion instead of Fast-dLLM on the diffusion modelDream-7B-Base · AlpacaEvalLC win rate: 40.2 -> 44.6 with 22.4x speedupImproved
use planned diffusion instead of autoregressive decodingDream-7B-Base · AlpacaEvalspeedup / LC win rate: 1.27x speedup, win rate 50.0 -> 49.2Mixed
use block-sparse attention between chunks (PDSA)Planned diffusion · AlpacaEvalspeedup / LC win rate: speedup 1.81x vs AR, win rate 44.6Mixed
increase fine-tuning epochs from 2 to 16 (ablation)Planned diffusion · AlpacaEvalLC win rate: 44.9 -> 49.2Improved
increase fine-tuning epochs from 2 to 16 (ablation)Autoregressive · AlpacaEvalLC win rate: flat at 50.0No clear effect
increase fine-tuning epochs from 2 to 16 (ablation)Diffusion · AlpacaEvalLC win rate: 41.1 -> 52.6Improved
increase fine-tuning epochs from 2 to 16 (ablation)PDSA · AlpacaEvalLC win rate: 40.2 -> 43.7Improved
include topic attribute in the plan (ablation)Planned diffusion (4 epochs) · AlpacaEvalLC win rate: 23.33 -> 46.65Improved
include <sync/> tokens in the plan (ablation)Planned diffusion (4 epochs) · AlpacaEvalLC win rate and latency: without sync: win rate 46.65 -> 42.96, latency 3.23 s -> 5.81 sMixed
scale predicted chunk length away from 1.0 (0.5 to 2.5) (ablation)Planned diffusion · AlpacaEvalLC win rateHurt
lower step ratio and confidence threshold (ablation)Planned diffusion · AlpacaEvalquality-latency trade-offMixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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