AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Mitigating the Effects of Non-Identifiability on Inference for Bayesian Neural Networks with Latent Variables

Yaniv Yacoby, Weiwei Pan, Finale Doshi-Velez

ICML 2023 · Probabilistic and scientific modeling › Causal discovery · approach: Latent-variable causal estimation

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
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Heavy Tailtest log-likelihood: -1.867 ± 0.078 -> -1.426 ± 0.042Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Goldbergtest log-likelihood: -1.026 ± 0.056 -> -0.963 ± 0.041Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Williamstest log-likelihood: -1.033 ± 0.156 -> -0.414 ± 0.184Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Yuantest log-likelihood: -1.278 ± 0.164 -> -1.211 ± 0.083Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Depewegtest log-likelihood: -2.342 ± 0.048 -> -1.973 ± 0.049Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Abalonetest log-likelihood: -0.843 ± 0.071 -> -0.831 ± 0.086Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Airfoiltest log-likelihood: -0.512 ± 0.083 -> -0.462 ± 0.056Improved
Use NCAI instead of mean-field VIBNN+LV · Energy Efficiencytest log-likelihood: 0.573 ± 0.288 -> 0.898 ± 0.452Hurt
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Lidartest log-likelihood: 0.129 ± 0.131 -> 0.263 ± 0.11Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Wine Quality Redtest log-likelihood: -1.709 ± 0.22 -> -0.849 ± 0.038Improved
Constrain latent noise to be less dependent on inputs and closer to its priorBNN+LV · Yachttest log-likelihood: 0.638 ± 0.121 -> 0.832 ± 0.077Improved
Select nonzero constraint strengths rather than fixing them at zero (ablation)BNN+LV NCAI · Heavy Tailtest log-likelihood: -1.481 ± 0.018 -> -1.426 ± 0.042Improved

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