AI Paper Atlas

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

Characteristic Circuits

Zhongjie Yu, Martin Trapp, Kristian Kersting

NeurIPS 2023 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Exact inference for probabilistic programs · 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 RDC-based structure learning instead of G-test-based structure learning (ablation)Characteristic circuits with α-stable leaves · Heterogeneous UCI datasetstest log-likelihoodMixed
Use α-stable rather than normal continuous leaves (ablation)Characteristic circuits · Heterogeneous UCI datasetstest log-likelihoodImproved
Model discrete variables as normal leaves rather than categorical leaves (ablation)CC-N · MM synthetic dataCFD: 0.0270 vs 0.0006Hurt
Learn parameters by minimizing CFD (ablation)Random-structure characteristic circuits · MM synthetic datatest log-likelihood: -4.93 -> -3.50Improved
Learn parameters by minimizing CFD (ablation)Random-structure characteristic circuits · BN synthetic datatest log-likelihood: -6.30 -> -4.12Improved
Learn the circuit structure from data instead of using a random structure, then learn parameters (ablation)Characteristic circuits · MM synthetic datatest log-likelihood: -3.50 -> -2.86Improved
Learn the circuit structure from data instead of using a random structure, then learn parameters (ablation)Characteristic circuits · BN synthetic datatest log-likelihood: -4.12 -> -3.27Improved
Use randomized weights and leaf parameters rather than structure-learning initialization (ablation)Structure-learned characteristic circuits with parameter learning · MM synthetic datatest log-likelihood: -2.86 -> -3.34Hurt
Use randomized weights and leaf parameters rather than structure-learning initialization (ablation)Structure-learned characteristic circuits with parameter learning · BN synthetic datatest log-likelihood: -3.27 -> -3.93Hurt
Use a structured characteristic circuit instead of a single empirical characteristic functionCharacteristic circuit with ECF leaves · MM and BN synthetic dataCFDImproved
Use RDC-based rather than G-test-based product-node splitting (ablation)Structure-learned characteristic circuits · Heterogeneous UCI datasetstest log-likelihoodImproved

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