analiza aiqbench
This research introduces Positional Task Conditioning (PTC), a method that decomposes complex product catalog error detection into focused sub-tasks to overcome LLM context limitations. By distilling capabilities into smaller models, the approach achieves near-frontier performance at significantly lower cost while scaling to millions of product families globally. The technique represents a practical advancement in applying machine learning to large-scale data consistency problems without relying solely on massive context windows.
Positional task conditioning for scalable defect detection across product families in large product catalogs
arXiv:2609.09567v1 Announce Type: new Abstract: Product families in large product catalogs suffer from inconsistencies such as duplicates and unit mismatches that degrade customer experience. Detecting these requires reasoning over multiple error types across lengthy product listings, where LLM classification quality degrades due to long-context limitations. We address this by decomposing detection into focused sub-tasks that reduce context and isolate error types, improving F1 from 52\% to 87\%. For scalable deployment, we introduce Positional Task Conditioning (PTC), which distills this capability into a single smaller model by reinforcing task identity at structural prompt boundaries. PTC outperforms rationale-based distillation across five models and two architecture families, achieving within 1.79\% F1 of the frontier at upto 98\% lower cost. Our system is deployed across multiple countries processing 10+ million product families.
Abstract:Product families in large product catalogs suffer from inconsistencies such as duplicates and unit mismatches that degrade customer experience. Detecting these requires reasoning over multiple error types across lengthy product listings, where LLM classification quality degrades due to long-context limitations. We address this by decomposing detection into focused sub-tasks that reduce context and isolate error types, improving F1 from 52\% to 87\%. For scalable deployment, we introduce Positional Task Conditioning (PTC), which distills this capability into a single smaller model by reinforcing task identity at structural prompt boundaries. PTC outperforms rationale-based distillation across five models and two architecture families, achieving within 1.79\% F1 of the frontier at upto 98\% lower cost. Our system is deployed across multiple countries processing 10+ million product families.
Machine Learning (cs.LG)
arXiv:2609.09567 [cs.LG]
(or arXiv:2609.09567v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.09567 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 · 217 cuvinte
