Jasper Snoek, Rodolphe Jenatton, Stanislav Fort, Balaji Lakshminarayanan, Dustin Tran, Marton Havasi, Jeremiah Liu, Andrew Dai
We lifted 5 functions out of this paper's own repositories and ran 4 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| ensta-u2is/torch-uncertainty | — | 4 of 5 |
| Function | Status | Where it lives |
|---|---|---|
| Mask1d | Ran | ensta-u2is/torch-uncertainty/src/torch_uncertainty/layers/masksembles.py code served (permissive licence) · get_code("8d8d122cf73afa3e") |
| _generate_masks | Ran | ensta-u2is/torch-uncertainty/src/torch_uncertainty/layers/masksembles.py code served (permissive licence) · get_code("6d8a981a433b52d5") |
| generate_masks | Ran | ensta-u2is/torch-uncertainty/src/torch_uncertainty/layers/masksembles.py code served (permissive licence) · get_code("5ca5c0443875a99b") |
| generation_wrapper | Ran | ensta-u2is/torch-uncertainty/src/torch_uncertainty/layers/masksembles.py code served (permissive licence) · get_code("11e6d40e826ec772") |
| MaskedLinear | Not yet run | ensta-u2is/torch-uncertainty/src/torch_uncertainty/layers/masksembles.py code served (permissive licence) · get_code("10cca0803fd92522") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Recent approaches to efficiently ensemble neural networks have shown that strong robustness and uncertainty performance can be achieved with a negligible gain in parameters over the original network. However, these methods still require multiple forward passes for prediction, leading to a significant computational cost. In this work, we show a surprising result: the benefits of using multiple predictions can be achieved 'for free' under a single model's forward pass. In particular, we show that, using a multi-input multi-output (MIMO) configuration, one can utilize a single model's capacity to train multiple subnetworks that independently learn the task at hand. By ensembling the predictions made by the subnetworks, we improve model robustness without increasing compute. We observe a significant improvement in negative log-likelihood, accuracy, and calibration error on CIFAR10, CIFAR100, ImageNet, and their out-of-distribution variants compared to previous methods.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2010.06610")
get_code_for_paper("2010.06610")
have("2010.06610")
Connect an agent — have() is free.