AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Episodic Multi-Task Learning with Heterogeneous Neural Processes

Jiayi Shen, Xiantong Zhen, Cheems Wang, Marcel Worring

NeurIPS 2023 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Meta-augmentation and meta-objectives · 4 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
remove both global latent z and local latent parameters w (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 67.09 -> 76.29Improved
add global latent representation z (vs. w only) (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 73.29 -> 76.29Improved
add local latent parameters w (vs. z only) (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 71.11 -> 76.29Improved
replace probabilistic transformer inference with MLP (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 76.29 -> 73.30Hurt
remove learnable tokens from transformer inference (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 76.29 -> 75.25Hurt
use deterministic instead of probabilistic inference modules (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 76.29 -> 70.61Hurt
remove learnable tokens, deterministic setting (ablation)HNPs deterministic · Office-Home 4-task 5-way 1-shotaverage accuracy: 70.22 -> 70.61Improved
generate w by concatenating z with context features (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 76.29 -> 69.74Hurt
generate w by adding z to context features (ablation)HNPs · Office-Home 4-task 5-way 1-shotaverage accuracy: 76.29 -> 72.87Hurt
increase shots from 1 to 20 (HNPs vs TNP-D gap)HNPs vs TNP-D · Office-Home 4-task 5-wayaverage accuracy: 1-shot 76.29 vs 65.49; 20-shot 81.56 vs 81.12Mixed
increase number of classes from 20 to 40HNPs vs TNP-D · DomainNet 6-task 1-shotaverage accuracy: 40-way 29.31 vs 21.64Mixed
vary number of Monte Carlo samples from 1 to 30HNPs · Office-Home 4-task 5-way 1-shotaverage accuracyNo 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