Alexandre Lacoste, Pau Rodríguez, Massimo Caccia, Laurent Charlin, Udem, Issam Laradji, David Vazquez, Element Ai, Lee Zamparo, Hec Montreal
We lifted 10 functions out of this paper's own repositories and ran 6 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 |
|---|---|---|
| sobieskibj/rcsb | — | 5 of 9 |
| issamlaradji/ssr | — | 1 of 1 |
| Function | Status | Where it lives |
|---|---|---|
| _cumulative_sum_threshold | Ran | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("c4ba5f856f01c3c0") |
| _get_last_conv_layer | Ran | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("5c104de03088fbae") |
| _normalize_scale | Ran | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("f3a716fb8b80641a") |
| cons_features_loss | Ran | issamlaradji/ssr/src/losses.py code served (permissive licence) · get_code("88def237af56a05a") |
| expand_from_cell_view | Ran | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("cf4c30b4c9dde913") |
| sum_into_cell_view | Ran | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("f123902550a574d0") |
| AttributionMethod | Not yet run | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("89a32ba43743ecf6") |
| ClassifierBase | Not yet run | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("73d849a584e707b8") |
| _get_layers_v2 | Not yet run | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("b2063be99abaab6c") |
| normalize_attr | Not yet run | sobieskibj/rcsb/src/explainers/base.py pointer only (licence: NONE) · get_code("f7b412e61259110d") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Explainability for machine learning models has gained considerable attention within the research community given the importance of deploying more reliable machinelearning systems. In computer vision applications, generative counterfactual methods indicate how to perturb a model's input to change its prediction, providing details about the model's decision-making. Current methods tend to generate trivial counterfactuals about a model's decisions, as they often suggest to exaggerate or remove the presence of the attribute being classified. For the machine learning practitioner, these types of counterfactuals offer little value, since they provide no new information about undesired model or data biases. In this work, we identify the problem of trivial counterfactual generation and we propose DiVE to alleviate it. DiVE learns a perturbation in a disentangled latent space that is constrained using a diversity-enforcing loss to uncover multiple valuable explanations about the model's prediction. Further, we introduce a mechanism to prevent the model from producing trivial explanations. Experiments on CelebA and Synbols demonstrate that our model improves the success rate of producing high-quality valuable explanations when compared to previous state-of-the-art methods. Code is available at https://github.com/ElementAI/ beyond-trivial-explanations.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2103.10226")
get_code_for_paper("2103.10226")
have("2103.10226")
Connect an agent — have() is free.