Rectified Lagrangian for Out-of-Distribution Detection in Modern Hopfield Networks
Authors: Ryo Moriai, Nakamasa Inoue, Masayuki Tanaka, Rei Kawakami, Satoshi Ikehata, Ikuro Sato
AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate the effectiveness of Rec Lag-based MHNs compared to energy-based OOD detection methods, including those using state-of-the-art Hopfield energies, across nine image datasets. |
| Researcher Affiliation | Collaboration | Ryo Moriai*1, Nakamasa Inoue*1, Masayuki Tanaka1, Rei Kawakami1, Satoshi Ikehata1, 3, Ikuro Sato1,2 1 Institute of Science Tokyo, Japan 2 Denso IT Laboratory, Inc. Japan 3 National Institute of Informatics, Japan EMAIL, EMAIL |
| Pseudocode | No | The paper includes mathematical equations and theoretical proofs (e.g., Theorem 1, Theorem 2, Theorem 3, Sketch of proof, full proofs in Appendix A, B, C), but no explicitly labeled pseudocode or algorithm blocks are present in the main text. |
| Open Source Code | No | The paper does not contain any explicit statement about releasing source code or provide a link to a code repository for the methodology described. |
| Open Datasets | Yes | Eleven image datasets were used to conduct OOD detection experiments: CIFAR-10 (Krizhevsky, Hinton et al. 2009), CIFAR-100 (Krizhevsky, Hinton et al. 2009), SVHN (Netzer et al. 2011), LSUN-C (Yu et al. 2015), LSUN-R (Yu et al. 2015), i SUN (Xu et al. 2015), Places365 (Zhou et al. 2017), DTD (Cimpoi et al. 2014), Tiny Image Net (TIN) (Deng et al. 2009), SUN (Xiao et al. 2010), and i Naturalist (Van Horn et al. 2018). |
| Dataset Splits | Yes | The CIFAR-10 or CIFAR-100 dataset was used as the ID dataset, and the other nine datasets were used as OOD datasets. ... They were trained on an ID dataset using cross-entropy loss. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU models, CPU types, or memory specifications used for running the experiments. It states, "Other implementation details are given in Appendix D.", but Appendix D is not provided in the main text. |
| Software Dependencies | No | The paper does not provide specific software dependencies or library versions used for the experiments. It states, "Other implementation details are given in Appendix D.", but Appendix D is not provided in the main text. |
| Experiment Setup | No | The paper mentions training on an ID dataset using cross-entropy loss, but it does not specify concrete hyperparameter values such as learning rate, batch size, number of epochs, or optimizer settings in the main text. It states, "Other implementation details are given in Appendix D.", but Appendix D is not provided in the main text. |