Supervised Learning Has a Geometric Blind Spot
Quick summary
arXiv:2604.21395v3 Announce Type: replace-cross Abstract: Ordinary supervised training minimises the task loss and then stops. It never pays for how far the representation moves when the input is nudged along directions that helped fit training labels---including directions that are nuisance at deployment. We call that leftover sensitivity the geometric blind spot of empirical risk minimisation. In a Gaussian linear model where the nuisance enters the label conditional and the decoder has finite Lipschitz constant, population MSE forces a floor on linearised representation drift. The same dist
Key takeaways
- arXiv:2604.21395v3 Announce Type: replace-cross Abstract: Ordinary supervised training minimises the task loss and then stops.
- It never pays for how far the representation moves when the input is nudged along directions that helped fit training labels---including directions that are nuisance at deployment.
- We call that leftover sensitivity the geometric blind spot of empirical risk minimisation.
Why it matters
“Supervised Learning Has a Geometric Blind Spot” should be evaluated beyond branding and benchmark scores. Its practical importance will emerge in task accuracy, latency, unit cost, safety and integration with real workflows.
Member comments