Residual Pathway Priors for Soft Equivariance Constraints
Authors: Marc Finzi, Gregory Benton, Andrew G. Wilson
NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The results for these 3 settings are given in Figure 4. Across all settings RPP-EMLP match the performance of EMLP when symmetries are exact, perform as well as an MLP when the symmetry is misspecified and better than both when the symmetry is approximate. and We evaluate RPPs on the standard suite of Mujoco continuous control tasks in the context of model-free reinforcement learning. |
| Researcher Affiliation | Academia | Marc Finzi New York University Greg Benton New York University Andrew Gordon Wilson New York University |
| Pseudocode | No | The paper does not contain any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | We provide a Py Torch implementation of residual pathway priors at https://github.com/mfinzi/residual-pathway-priors. |
| Open Datasets | Yes | Using RPP-Conv specified by the prior in Eqn 1 we apply the model to CIFAR-10 classification and UCI regression tasks where the inputs are reshaped to zero-padded two dimensional arrays and treated as images. and [11] Dheeru Dua and Casey Graff. UCI machine learning repository, 2017. URL http://archive. ics.uci.edu/ml. |
| Dataset Splits | No | The full details for the architectures and training procedure are given in Appendix C. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU/CPU models or specific cloud computing resources used for running experiments. |
| Software Dependencies | No | The paper mentions a 'PyTorch implementation' but does not specify the version number for PyTorch or any other software dependencies. |
| Experiment Setup | Yes | For these experiments we use a prior variance of σ2 a = 105 on the EMLP weights and σ2 b = 1 on the MLP weights. |