Enhancing Neural Subset Selection: Integrating Background Information into Set Representations
Authors: Binghui Xie, Yatao Bian, Kaiwen Zhou, Yongqiang Chen, Peilin Zhao, Bo Han, Wei Meng, James Cheng
ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Comprehensive empirical evaluations across diverse tasks and datasets validate the enhanced efficacy of our approach over conventional methods, underscoring the practicality and potency of our proposed strategies in real-world contexts. |
| Researcher Affiliation | Collaboration | Binghui Xie1, Yatao Bian2, Kaiwen zhou1, Yongqiang Chen1 1The Chinese University of Hong Kong, 2Tencent AI Lab, 3Hong Kong Baptist University |
| Pseudocode | No | The paper describes algorithms and implementations verbally and with equations, but it does not include any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide an explicit statement or a link indicating the availability of its source code. |
| Open Datasets | Yes | We use the dataset (Gillenwater et al., 2014a) from the Amazon baby registry for this experiment...We conduct set anomaly detection tasks on three real-world datasets: the double MNIST (Sun, 2019), the Celeb A (Liu et al., 2015b) and the F-MNIST (Xiao et al., 2017)...We conduct experiments using the following datasets: PDBBind (Liu et al., 2015a) and Binding DB (Liu et al., 2007). |
| Dataset Splits | Yes | Each dataset is divided into the training, validation, and test sets with sizes of 10,000, 1,000, and 1,000, respectively...Then we divided the remaining subsets into the training, validation, and test folds with a 1 : 1 : 1 ratio. |
| Hardware Specification | Yes | Notably, we choose the largest batch size that allows the model to be trained on a single Ge Force RTX 2080 Ti GPU, ensuring efficient training. |
| Software Dependencies | No | The paper mentions software components like 'Adam optimizer (Kingma & Ba, 2014)', 'Res Net', 'BERT model', 'ACNN', and 'Deep DTA model' but does not specify their version numbers or the versions of general programming languages/libraries like Python or PyTorch. |
| Experiment Setup | Yes | The proposed models are trained using the Adam optimizer (Kingma & Ba, 2014) with a fixed learning rate of 1e 4 and a weight decay rate of 1e 5. To accommodate the varying model sizes across different datasets, we select the batch size from the set {4, 8, 16, 32, 64, 128}. |