Set-based Meta-Interpolation for Few-Task Meta-Learning
Authors: Seanie Lee, Bruno Andreis, Kenji Kawaguchi, Juho Lee, Sung Ju Hwang
NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We empirically validate the efficacy of Meta-Interpolation on eight datasets spanning across various domains such as image classification, molecule property prediction, text classification and sound classificattion. Experimentally, we show that Meta-Interpolation consistently outperforms all the relevant baselines. |
| Researcher Affiliation | Collaboration | KAIST1, National University of Singapore2, AITRICS3 |
| Pseudocode | Yes | Algorithm 1 Meta-training; Algorithm 2 Hyper Grad [27] |
| Open Source Code | No | The paper states in its self-assessment that code is included or available in supplementary material/URL, but does not provide a direct link or explicit statement within the main paper text detailing where to access the code for their methodology. |
| Open Datasets | Yes | (1), (2), & (3) Metabolism [17], NCI [31] and Tox21 [18]: these are binary classification datasets for predicting the properties of chemical molecules. (4) GLUE-Sci Tail [30]: it consists of four natural language inference datasets... (5) ESC-50 [34]: this is an environmental sound recognition dataset. (6) Rainbow MNIST (RMNIST) [11]: this is a 10-way classification dataset. (7) & (8) Mini-Image Net-S [45] and CIFAR100-FS [22]: these are 5-way classification datasets... |
| Dataset Splits | Yes | For Metabolism, we use three subdatasets for meta-training, meta-validation, and meta-testing, respectively. For NCI, we use four subdatasets for meta-training, two for meta-validation and the remaining three for meta-testing. For Tox21, we use six subdatasets for meta-training, two for meta-validation, and four for meta-testing. We make a 20/15/15 split out of 50 base classes for meta-training/validation/testing and sample 5 classes from each spilt to construct a 5-way classification task. For Mini-Image Net-S and CIFAR100-FS, following Yao et al. [50], we choose 12/16/20 classes out of 100 base classes for meta-training/validation/testing, respectively and sample 5 classes from each split to construct a task. |
| Hardware Specification | No | The paper does not provide specific details about the hardware used for running the experiments (e.g., GPU model, CPU type, memory). |
| Software Dependencies | No | The paper mentions software like RDKit, ELECTRA, and VGGish, but it does not provide specific version numbers for these or any other software dependencies. |
| Experiment Setup | Yes | For RMNIST, Mini-Image Net-S, and CIFAR100-FS, we use four convolutional blocks with each block consisting of a convolution, ReLU, batch normalization [19], and max pooling. For Metabolism, Tox21, and NCI, we convert the chemical molecules into SMILES format and extract a 1024 bit fingerprint feature using RDKit [15]... We use two blocks of affine transformation, batch normalization, and Leaky ReLU, and affine transformation for the last layer. For GLUE-Sci Tail dataset, we stack 3 fully connected layers with ReLU on the pretrained language model ELECTRA [8]. For ESC-50 dataset, we pass raw audio signal to the pretrained VGGish [16] feature extractor to obtain an embedding vector. We use the feature vector as input to the classifier which is exactly the same as the one used for Metabolism, Tox21, and NCI. For our Meta-Interpolation, we use Set Transformer [23] for the set function φλ. Require: Tasks {T train t }T t=1 {T val t}T t =1, learning rate α, η R+, update period S, and batch size B. |