SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2409.16756 · NeurIPS · 2024

Navigating the Maze of Explainable AI: A Systematic Approach to Evaluating Methods and Metrics

Udo Schlegel, Mennatallah El-Assady, Carsten Lüth, Till Bungert, Lukas Klein, Paul Jäger

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 11 functions out of this paper's own repositories and ran 5 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.

RepositoryRoleRan
iml-dkfz/latec canonical 3 of 8
IML-DKFZ/latec canonical 2 of 3
FunctionStatusWhere it lives
feature_mask Ran iml-dkfz/latec/src/utils/reshape_transforms.py
code served (permissive licence) · get_code("aba56ac24bacff5a")
get_hidden_layer_eval Ran iml-dkfz/latec/src/utils/hidden_layer_selection.py
code served (permissive licence) · get_code("959264dd199e81b7")
get_pylogger Ran iml-dkfz/latec/src/utils/pylogger.py
code served (permissive licence) · get_code("aebb727eed75bd10")
load_evaluation_scores Ran IML-DKFZ/latec/src/main/main_rank.py
code served (permissive licence) · get_code("68054fd89f9d291d")
normalize_data Ran IML-DKFZ/latec/src/main/main_rank.py
code served (permissive licence) · get_code("61272ae2b64a89b2")
add_p_value_annotation Not yet run iml-dkfz/latec/src/utils/plot_utils.py
code served (permissive licence) · get_code("f5d0af3a9faa7bc7")
get_hidden_layer Not yet run iml-dkfz/latec/src/utils/hidden_layer_selection.py
code served (permissive licence) · get_code("6da1743ee500f0fb")
left_align_facet_plot_titles Not yet run iml-dkfz/latec/src/utils/plot_utils.py
code served (permissive licence) · get_code("2793fcdf7c5b1c54")
prepare_modalities Not yet run IML-DKFZ/latec/src/main/main_rank.py
code served (permissive licence) · get_code("c8cd26a07b4abbd3")
reshape_transform_2D Not yet run iml-dkfz/latec/src/utils/reshape_transforms.py
code served (permissive licence) · get_code("361997639f735d7a")
reshape_transform_3D Not yet run iml-dkfz/latec/src/utils/reshape_transforms.py
code served (permissive licence) · get_code("22c188f68d075a1a")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Explainable AI (XAI) is a rapidly growing domain with a myriad of proposed methods as well as metrics aiming to evaluate their efficacy. However, current studies are often of limited scope, examining only a handful of XAI methods and ignoring underlying design parameters for performance, such as the model architecture or the nature of input data. Moreover, they often rely on one or a few metrics and neglect thorough validation, increasing the risk of selection bias and ignoring discrepancies among metrics. These shortcomings leave practitioners confused about which method to choose for their problem. In response, we introduce LATEC, a large-scale benchmark that critically evaluates 17 prominent XAI methods using 20 distinct metrics. We systematically incorporate vital design parameters like varied architectures and diverse input modalities, resulting in 7,560 examined combinations. Through LATEC, we showcase the high risk of conflicting metrics leading to unreliable rankings and consequently propose a more robust evaluation scheme. Further, we comprehensively evaluate various XAI methods to assist practitioners in selecting appropriate methods aligning with their needs. Curiously, the emerging top-performing method, Expected Gradients, is not examined in any relevant related study. LATEC reinforces its role in future XAI research by publicly releasing all 326k saliency maps and 378k metric scores as a (meta-)evaluation dataset. The benchmark is hosted at: https://github.com/IML-DKFZ/latec.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2409.16756")
get_code_for_paper("2409.16756")
have("2409.16756")

Connect an agent — have() is free.