ANML: Attribution-Native Machine Learning with Guaranteed Robustness

TL;DR

ANML introduces a novel training framework that weights data samples based on multiple quality factors, enhancing model robustness, performance, and attribution capabilities across diverse datasets.

Contribution

It proposes a new attribution-native training method that combines gradient signals with data provenance to improve robustness and enable contributor attribution.

Findings

Achieves 33-72% error reduction over gradient-only baselines.

Data-efficient: 20% high-quality data outperforms 100% uniform data by 47%.

Contributor attribution outperforms sample-level detection under subtle corruption.

Abstract

Frontier AI systems increasingly train on specialized expert data, from clinical records to proprietary research to curated datasets, yet current training pipelines treat all samples identically. A Nobel laureate's contribution receives the same weight as an unverified submission. We introduce ANML (Attribution-Native Machine Learning), a framework that weights training samples by four quality factors: gradient-based consistency (q), verification status (v), contributor reputation (r), and temporal relevance (T). By combining what the model observes (gradient signals) with what the system knows about data provenance (external signals), ANML produces per-contributor quality weights that simultaneously improve model performance and enable downstream attribution. Across 5 datasets (178-32,561 samples), ANML achieves 33-72% error reduction over gradient-only baselines. Quality-weighted training is data-efficient: 20% high-quality data outperforms 100% uniformly weighted data by 47%. A Two-Stage Adaptive gating mechanism guarantees that ANML never underperforms the best available baseline, including under strategic joint attacks combining credential faking with gradient alignment. When per-sample detection fails against subtle corruption, contributor-level attribution provides 1.3-5.3x greater improvement than sample-level methods, with the advantage growing as corruption becomes harder to detect.