Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
RNNLogic: Learning Logic Rules for Reasoning on Knowledge Graphs
Authors: Meng Qu, Junkun Chen, Louis-Pascal Xhonneux, Yoshua Bengio, Jian Tang
ICLR 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on four datasets prove the effectiveness of RNNLogic. |
| Researcher Affiliation | Academia | 1Mila Qu ebec AI Institute 2Universit e de Montr eal 3Tsinghua University 4HEC Montr eal 5Canadian Institute for Advanced Research (CIFAR) |
| Pseudocode | Yes | Algorithm 1 Workflow of RNNLogic |
| Open Source Code | Yes | The codes of RNNLogic are available: https://github.com/Deep Graph Learning/RNNLogic |
| Open Datasets | Yes | We choose four datasets for evaluation, including FB15k-237 (Toutanova & Chen, 2015), WN18RR (Dettmers et al., 2018), Kinship and UMLS (Kok & Domingos, 2007). |
| Dataset Splits | Yes | For Kinship and UMLS, there are no standard data splits, so we randomly sample 30% of all the triplets for training, 20% for validation, and the rest 50% for testing. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU or CPU models, memory, or specific cloud computing instance types used for experiments. |
| Software Dependencies | No | The paper mentions software components like 'LSTM' and 'Adam optimizer' but does not provide specific version numbers for any libraries, frameworks, or programming languages. |
| Experiment Setup | Yes | For the rule generator, the maximum length of generated rules is set to 4 for FB15k-237, 5 for WN18RR, and 3 for the rest... The size of input and hidden states in RNNθ are set to 512 and 256. The learning rate is set to 1 10 3 and monotonically decreased in a cosine shape. |