Environment Diversification with Multi-head Neural Network for Invariant Learning

Authors: Bo-Wei Huang, Keng-Te Liao, Chang-Sheng Kao, Shou-De Lin

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 4 Experiments We empirically validate the proposed method on biased datasets, Adult-Confounded, CMNIST, Waterbirds and SNLI.
Researcher Affiliation Academia Bo-Wei Huang Keng-Te Liao Chang-Sheng Kao Shou-De Lin Department of Computer Science and Information Engineering National Taiwan University, Taiwan {r10922007,d05922001,b07902046,sdlin}@csie.ntu.edu.tw
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks clearly labeled as such.
Open Source Code No The paper does not provide any concrete access to source code, such as a specific repository link or an explicit code release statement, for the methodology described.
Open Datasets Yes We empirically validate the proposed method on biased datasets, Adult-Confounded, CMNIST, Waterbirds and SNLI. ... We take UCI Adult [16] ... We report our evaluation on a noisy digit recognition dataset, CMNIST. Following [1], we first assign Y = 1 to those whose digits are smaller than 5 and Y = 0 to the others. ... In Waterbirds [29], each bird photograph, from CUB dataset [31], is combined with one background image, from Places dataset [33]. ... The target of SNLI [3] is to predict the relation between two given sentences, premise and hypothesis.
Dataset Splits Yes For hyper-parameter tuning, we split 10% of training data to construct an in-distribution validation set. ... To split a validation set whose overall distribution is i.i.d. to the training set, we merge original training and validation data 3 and split 10% for hyper-parameter tuning.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or memory amounts) used for running its experiments.
Software Dependencies No The paper mentions software like Resnet, Distil BERT, and other models/frameworks, but it does not specify version numbers for any key software components or libraries required to replicate the experiment.
Experiment Setup Yes For hyper-parameter tuning, we split 10% of training data to construct an in-distribution validation set. ... For all competitors, MLP is taken as the base model and full-batch training is implemented. ... MLP with one hidden layer of 96 neurons is considered. ... MLP with two hidden layers of 390 neurons... Resnet-34 [14] is chosen for mini-batch fine-tuning. ... Distil BERT [30] is taken as the pre-trained model for further mini-batch fine-tuning.