AI Paper Atlas

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

Variational Inference with Tail-adaptive f-Divergence

Dilin Wang, Hao Liu, Qiang Liu

NeurIPS 2018 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Variational inference theory and gradients · 17 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 tail-adaptive f-divergence (β=-1.0) instead of α=0.0Bayesian neural network regression · Bostontest RMSE: 2.956±0.171 -> 2.828±0.177Improved
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.5Bayesian neural network regression · Concretetest RMSE: 5.381±0.111 -> 5.371±0.115No clear effect
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.0Bayesian neural network regression · Energytest log-likelihood: -1.795±0.029 -> -1.758±0.034Improved
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.5Bayesian neural network regression · Kin8nmtest RMSE: 0.083±0.001 -> 0.085±0.001Hurt
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.0Bayesian neural network regression · Navaltest RMSE: 0.001±0.000 -> 0.001±0.000No clear effect
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.5Bayesian neural network regression · Combinedtest RMSE: 4.154±0.042 -> 4.116±0.032Improved
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.0Bayesian neural network regression · Winetest RMSE: 0.634±0.007 -> 0.636±0.008Hurt
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.0Bayesian neural network regression · Yachttest log-likelihood: -1.751±0.056 -> -1.711±0.059Improved
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.0Bayesian neural network regression · Proteintest RMSE: 4.565±0.026 -> 4.487±0.019Improved
Use tail-adaptive f-divergence (β=-1.0) instead of α=0.5Bayesian neural network regression · Yeartest log-likelihood: -3.518±0.042 -> -3.570±0.037Hurt
Use tail-adaptive f-divergence instead of fixed α-divergenceSAC · MuJoCo benchmark tasksfinal average rewardImproved
Increase α from the original SAC setting α=0SAC · MuJoCo environmentsaverage rewardImproved

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