analiza aiqbench
This arXiv preprint proposes a pretraining method for transaction risk models that encodes high-cardinality identifiers (card, device, account) by their recurrence patterns within a sequence rather than their raw values. The reported gains, 4.13% on an industrial dataset and 8.83% on public benchmarks, come from the authors' own experiments and have not been independently validated. The claim that performance holds when all identity values are unseen during training is the most relevant point for readers working on generalization in production fraud systems.
ID2Pattern: Pretraining Generalizable Identity Behavior Patterns for Transaction Risk Foundation Models
arXiv:2610.41988v1 Announce Type: new Abstract: Large-scale behavioral pretraining offers a promising route to generalizable representations in online payment systems. However, high-cardinality identity fields such as card, device, and account IDs, though rich in behavioral signals, remain underexploited. Their continually growing vocabularies and limited semantic meaning make existing vocabulary-based and semantics-based encodings ill-suited. Their generalizable signal lies not in specific values, but in the identity behavior patterns they exhibit across a transaction sequence. Building on this insight, we propose ID2Pattern, a pretraining framework that brings these patterns into transaction foundation models. ID2Pattern first encodes identity fields with a Local Anonymous ID (LAID) tokenizer, which captures how each identity appears and recurs within the sequence, independent of its specific value. It then pretrains a multi-scale pattern codebook on unlabeled sequences, which abstracts LAID sequences into discrete behavior patterns for downstream prediction. Experiments on an industrial dataset with over 100M sequences and six public datasets show that ID2Pattern consistently outperforms strong baselines, improving industrial by 4.13% and public datasets by 8.83% on average. Notably, ID2Pattern remains equally effective even when all identity values are unseen during training, highlighting its potential to generalize to new users and markets.
Abstract:Large-scale behavioral pretraining offers a promising route to generalizable representations in online payment systems. However, high-cardinality identity fields such as card, device, and account IDs, though rich in behavioral signals, remain underexploited. Their continually growing vocabularies and limited semantic meaning make existing vocabulary-based and semantics-based encodings ill-suited. Their generalizable signal lies not in specific values, but in the identity behavior patterns they exhibit across a transaction sequence. Building on this insight, we propose ID2Pattern, a pretraining framework that brings these patterns into transaction foundation models. ID2Pattern first encodes identity fields with a Local Anonymous ID (LAID) tokenizer, which captures how each identity appears and recurs within the sequence, independent of its specific value. It then pretrains a multi-scale pattern codebook on unlabeled sequences, which abstracts LAID sequences into discrete behavior patterns for downstream prediction. Experiments on an industrial dataset with over 100M sequences and six public datasets show that ID2Pattern consistently outperforms strong baselines, improving industrial by 4.13% and public datasets by 8.83% on average. Notably, ID2Pattern remains equally effective even when all identity values are unseen during training, highlighting its potential to generalize to new users and markets.
Artificial Intelligence (cs.AI)
arXiv:2610.41988 [cs.AI]
(or arXiv:2610.41988v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2610.41988 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 283 cuvinte
Subiecte conexe
MosaicIMU: Composing Carrier Experts for Generalizable Neural Inertial Odometry
arXiv · 9 iun. 2026
Verifier-Guided Model Discovery for Physical Dynamical Systems with Pretrained Symbolic Transformers
arXiv · 5 aug. 2026
Multimodal Pretraining for Generalizable EEG Representation Learning
arXiv · 24 iul. 2026
MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image
arXiv · 30 sept. 2026