# Token Factory: Efficiently Integrating Diverse Signals into Large Recommendation Models

arXiv: https://arxiv.org/abs/2606.19635 · HTML: https://mlsanity.com/paper/2606.19635/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 27 references that are in the map.
- First posted on arXiv 17 Jun 2026.

## Builds on (most cited first, 11 of 11)

- [Continual Learning with Pre-Trained Models: A Survey](https://mlsanity.com/paper/2401.16386/) (2024)
- [Better Generalization with Semantic IDs: A Case Study in Ranking for Recommendations](https://mlsanity.com/paper/2306.08121/) (2023)
- [LONGER: Scaling Up Long Sequence Modeling in Industrial Recommenders](https://mlsanity.com/paper/2505.04421/) (2025)
- [RankMixer: Scaling Up Ranking Models in Industrial Recommenders](https://mlsanity.com/paper/2507.15551/) (2025)
- [TokenRec: Learning to Tokenize ID for LLM-based Generative Recommendation](https://mlsanity.com/paper/2406.10450/) (2024)
- [Collaboration of Large Language Models and Small Recommendation Models for Device-Cloud Recommendation](https://mlsanity.com/paper/2501.05647/) (2025)
- [Generative Retrieval with Semantic Tree-Structured Item Identifiers via Contrastive Learning](https://mlsanity.com/paper/2309.13375/) (2023)
- [From IDs to Semantics: A Generative Framework for Cross-Domain Recommendation with Adaptive Semantic Tokenization](https://mlsanity.com/paper/2511.08006/) (2025)
- [GenRec: A Preference-Oriented Generative Framework for Large-Scale Recommendation](https://mlsanity.com/paper/2604.14878/) (2026)
- [HyMiRec: A Hybrid Multi-interest Learning Framework for LLM-based Sequential Recommendation](https://mlsanity.com/paper/2510.13738/) (2025)
- [TokenMixer-Large: Scaling Up Large Ranking Models in Industrial Recommenders](https://mlsanity.com/paper/2602.06563/) (2026)

## Built on by (most cited first, 1 of 1)

- [KuaFu: Compressing Long User Behavior into Understanding at Billion Scale](https://mlsanity.com/paper/2609.31045/) (2026)

## Abstract (arXiv)

Large Recommendation Models (LRMs) have demonstrated promising capabilities in industry-scale recommendation tasks. However, holistically integrating traditional signals into these transformer-based architectures effectively and efficiently remains a major challenge. Conventional approaches that "textualize" these signals directly or create discrete item representations often lead to excessively long prompts, substantial memory footprints, and high computational overhead. To overcome these limitations, we propose "Token Factory", a framework designed to transform traditional signals into "soft tokens" that can be directly processed by LRMs. This approach enables efficient integration and compression of heterogeneous input features, preventing prompt length explosion while enhancing model performance. We detail the architecture of Token Factory and present experimental results validating its effectiveness in a production-scale recommendation environment.