AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Diamond Maps: Efficient Reward Alignment via Stochastic Flow Maps

Peter Holderrieth, Douglas Chen, Luca Eyring, Ishin Shah, Giri Anantharaman, Yutong He, Zeynep Akata, Tommi Jaakkola, Nicholas Matthew Boffi, Max Simchowitz

ICML 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Flow matching and flow sampling · 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 Monte Carlo particles (mc=64) to Diamond MapsWeighted Diamond Maps (FLUX Flow Map) · FLUX GenEvalGenEval: reach 0.785 at mc=64 vs Flow Map Guidance 0.737 and Best-of-N 0.748Improved
use Weighted Diamond Maps instead of Best-of-NFLUX Flow Map · UniGenBench++UniGenBench++ score: 0.637 -> 0.653Improved
use Weighted Diamond Maps instead of Flow Map GuidanceFLUX Flow Map · UniGenBench++UniGenBench++ score: 0.627 -> 0.653Improved
add more guidance steps to Flow Map Guidance (ablation)Flow Map Guidance · FLUX GenEvalGenEval: peak 0.737, after which performance declines with additional stepsHurt
use SNR=1.5 perturbation noise instead of SNR=1.1 (ablation)Weighted Diamond Maps · FLUX GenEvalGenEval: SNR=1.5 yields the strongest results, insufficient noise (SNR=1.1) limits explorationImproved
use excessive perturbation noise (SNR>=3.0) (ablation)Weighted Diamond Maps · FLUX GenEvalGenEval: SNR>=3.0 degrades quality, particularly with g=24Hurt
increase Monte Carlo samples mc (ablation)Weighted Diamond Maps · FLUX GenEvalGenEval: increasing mc yields consistent gainsImproved
use guidance frequency g=10 (ablation)Weighted Diamond Maps · FLUX GenEvalGenEval: g=10 offers the best trade-off between alignment strength and trajectory coherenceImproved
add weighted diamond-map guidance to SANA-Sprint 0.6BSANA-Sprint 0.6B · GenEvalGenEval mean: 0.70 -> 0.83Improved
add weighted diamond-map guidance to FLUX Flow MapFLUX Flow Map · GenEvalGenEval mean: 0.57 -> 0.79Improved
use Weighted Diamond Maps instead of ReNO on SANA-SprintSANA-Sprint 0.6B · GenEvalGenEval mean: 0.81 -> 0.83 with time 30.0 -> 10.0Improved
use Diamond DDPM early stop sampling instead of iterative denoising and noising (ablation)Posterior Diamond Maps · posterior samplingsample qualityImproved

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