AI Paper Atlas

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

On Parameter Tying by Quantization

Li Chou, Somdeb Sarkhel, Nicholas Ruozzi, Vibhav Gogate

AAAI 2016 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Annealing and divergence-based variational objectives · 8 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
relearn parameters after quantization (tied) versus quantization only (ablation)quantized MLE · bn2o networksmax parameter error: 0.339 -> 0.299Improved
relearn parameters after quantization versus quantization only (ablation)quantized MLE · students networkmax parameter error: 0.141 -> 0.140Improved
relearn parameters after quantization versus quantization only (ablation)quantized MLE · grid networkmax parameter error: 0.0186 -> 0.0181Improved
relearn parameters after quantization versus quantization only (ablation)quantized MLE · friends networkmax parameter error: 0.008 -> 0.007Improved
quantize MLE weights without relearningMLE · bn2o, students, grid, friendsmax parameter error: 0.339 -> 0.339No clear effect
quantization alone at low kMLE · noisy-or networkstest average log-likelihoodHurt
relearning after quantizationquantized model · noisy-or networkstest average log-likelihoodImproved
increase k toward true number of parametersquantized/relearned model · noisy-or networkslog-likelihood gap to MLEImproved
tied-weight slice importance sampling (TW) instead of MC-SAT and Gibbs on relearned parametersinference · students and grid networksHellinger distance of marginalsImproved
TW versus MC-SAT on untied MLE parametersinference · noisy-or, relational, grid networksHellinger distance of marginalsNo clear effect

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