AI Paper Atlas

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

Variational Annealing on Graphs for Combinatorial Optimization

Sebastian Sanokowski, Wilhelm Franz Berghammer, Sepp Hochreiter, Sebastian Lehner

NeurIPS 2023 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Exact inference for probabilistic programs · 1 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 annealed entropy regularization (T0 > 0 vs T0 = 0) (ablation)VAG-CO · MVC RB-100relative errorImproved
Add subgraph tokenization (k=5) vs none (ablation)VAG-CO · MVC RB-100relative errorImproved
Increase subgraph tokenization size k from 1 to 10 (ablation)VAG-CO · MVC RB-100AR*Improved
Use autoregressive VAG-CO instead of mean-field MFA-AnnealVAG-CO · MIS ENZYMESAR-hat: 0.968 -> 0.9863Improved
Use VAG-CO instead of MFA-AnnealVAG-CO · MIS COLLABAR-hat: 0.917 -> 0.993Improved
Use VAG-CO instead of MFA-AnnealVAG-CO · MaxCl ENZYMESAR-hat: 0.806 -> 0.943Improved
Use VAG-CO instead of MFA: CEVAG-CO · MIS ENZYMESAR*: 0.9875 -> 0.9960Improved
Use VAG-CO instead of MFA-Anneal: CEVAG-CO · MIS COLLABAR*: 0.9991 -> 0.9988No clear effect
Use VAG-CO instead of MFA: CEVAG-CO · MVC COLLABAR*: 1.00021 -> 1.0017Hurt
Use VAG-CO instead of MFA-Anneal: CEVAG-CO · MaxCl ENZYMESAR*: 0.9921 -> 0.987Hurt
Use VAG-CO instead of Meta-EGNVAG-CO · MVC TWITTERAR*: 1.019 -> 1.0024Improved
Use VAG-CO instead of GFlowNetVAG-CO · MaxCut BA 200-300average cut value: 704.30 -> 722.31Improved

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