Akshit Achara
@akshitra.bsky.social
📄 OSCAR: an auditing framework for localising shortcut reliance in pixel space.
We turn attribution maps into interpretable statistics to quantify spurious effects. Feedback welcome!
arxiv.org/abs/2512.18888
#ComputerVision #FairML #ExplainableAI #XAI #AI
We turn attribution maps into interpretable statistics to quantify spurious effects. Feedback welcome!
arxiv.org/abs/2512.18888
#ComputerVision #FairML #ExplainableAI #XAI #AI
Localising Shortcut Learning in Pixel Space via Ordinal Scoring Correlations for Attribution Representations (OSCAR)
Deep neural networks often exploit shortcuts. These are spurious cues which are associated with output labels in the training data but are unrelated to task semantics. When the shortcut features are a...
arxiv.org
December 27, 2025 at 11:48 PM
📄 OSCAR: an auditing framework for localising shortcut reliance in pixel space.
We turn attribution maps into interpretable statistics to quantify spurious effects. Feedback welcome!
arxiv.org/abs/2512.18888
#ComputerVision #FairML #ExplainableAI #XAI #AI
We turn attribution maps into interpretable statistics to quantify spurious effects. Feedback welcome!
arxiv.org/abs/2512.18888
#ComputerVision #FairML #ExplainableAI #XAI #AI
Reposted by Akshit Achara
Achara, Triantafillou, Puyol-Ant\'on, Hammers, King: Localising Shortcut Learning in Pixel Space via Ordinal Scoring Correlations for Attribution Representations (OSCAR) https://arxiv.org/abs/2512.18888 https://arxiv.org/pdf/2512.18888 https://arxiv.org/html/2512.18888
December 23, 2025 at 6:31 AM
Achara, Triantafillou, Puyol-Ant\'on, Hammers, King: Localising Shortcut Learning in Pixel Space via Ordinal Scoring Correlations for Attribution Representations (OSCAR) https://arxiv.org/abs/2512.18888 https://arxiv.org/pdf/2512.18888 https://arxiv.org/html/2512.18888