Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.17533 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917351478263808 |
|---|---|
| author | De Nadai, Marco D'Amico, Edoardo Lefarov, Max Tamborrino, Alexandre Vohra, Divita VanMiddlesworth, Mark Lin, Shawn Wood, Jacqueline Stypka, Jan Klyce, Eliza Dai, Keshi Heath, Timothy Christopher Gould, Martin D. Raimond, Yves Ghael, Sandeep Jebara, Tony Damianou, Andreas Radosavljevic, Vladan Bennett, Paul N. Lalmas, Mounia Chandar, Praveen |
| author_facet | De Nadai, Marco D'Amico, Edoardo Lefarov, Max Tamborrino, Alexandre Vohra, Divita VanMiddlesworth, Mark Lin, Shawn Wood, Jacqueline Stypka, Jan Klyce, Eliza Dai, Keshi Heath, Timothy Christopher Gould, Martin D. Raimond, Yves Ghael, Sandeep Jebara, Tony Damianou, Andreas Radosavljevic, Vladan Bennett, Paul N. Lalmas, Mounia Chandar, Praveen |
| contents | LLMs are increasingly applied to recommendation, retrieval, and reasoning, yet deploying a single end-to-end model that can jointly support these behaviors over large, heterogeneous catalogs remains challenging. Such systems must generate unambiguous references to real items, handle multiple entity types, and operate under strict latency and reliability constraints requirements that are difficult to satisfy with text-only generation. While tool-augmented recommender systems address parts of this problem, they introduce orchestration complexity and limit end-to-end optimization. We view this setting as an instance of a broader research problem: how to adapt LLMs to reason jointly over multiple-domain entities, users, and language in a fully self-contained manner. To this end, we introduce NEO, a framework that adapts a pre-trained decoder-only LLM into a tool-free, catalog-grounded generator. NEO represents items as SIDs and trains a single model to interleave natural language and typed item identifiers within a shared sequence. Text prompts control the task, target entity type, and output format (IDs, text, or mixed), while constrained decoding guarantees catalog-valid item generation without restricting free-form text. We refer to this instruction-conditioned controllability as language-steerability. We treat SIDs as a distinct modality and study design choices for integrating discrete entity representations into LLMs via staged alignment and instruction tuning. We evaluate NEO at scale on a real-world catalog of over 10M items across multiple media types and discovery tasks, including recommendation, search, and user understanding. In offline experiments, NEO consistently outperforms strong task-specific baselines and exhibits cross-task transfer, demonstrating a practical path toward consolidating large-scale discovery capabilities into a single language-steerable generative model. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_17533 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | A Unified Language Model for Large Scale Search, Recommendation, and Reasoning De Nadai, Marco D'Amico, Edoardo Lefarov, Max Tamborrino, Alexandre Vohra, Divita VanMiddlesworth, Mark Lin, Shawn Wood, Jacqueline Stypka, Jan Klyce, Eliza Dai, Keshi Heath, Timothy Christopher Gould, Martin D. Raimond, Yves Ghael, Sandeep Jebara, Tony Damianou, Andreas Radosavljevic, Vladan Bennett, Paul N. Lalmas, Mounia Chandar, Praveen Information Retrieval Machine Learning LLMs are increasingly applied to recommendation, retrieval, and reasoning, yet deploying a single end-to-end model that can jointly support these behaviors over large, heterogeneous catalogs remains challenging. Such systems must generate unambiguous references to real items, handle multiple entity types, and operate under strict latency and reliability constraints requirements that are difficult to satisfy with text-only generation. While tool-augmented recommender systems address parts of this problem, they introduce orchestration complexity and limit end-to-end optimization. We view this setting as an instance of a broader research problem: how to adapt LLMs to reason jointly over multiple-domain entities, users, and language in a fully self-contained manner. To this end, we introduce NEO, a framework that adapts a pre-trained decoder-only LLM into a tool-free, catalog-grounded generator. NEO represents items as SIDs and trains a single model to interleave natural language and typed item identifiers within a shared sequence. Text prompts control the task, target entity type, and output format (IDs, text, or mixed), while constrained decoding guarantees catalog-valid item generation without restricting free-form text. We refer to this instruction-conditioned controllability as language-steerability. We treat SIDs as a distinct modality and study design choices for integrating discrete entity representations into LLMs via staged alignment and instruction tuning. We evaluate NEO at scale on a real-world catalog of over 10M items across multiple media types and discovery tasks, including recommendation, search, and user understanding. In offline experiments, NEO consistently outperforms strong task-specific baselines and exhibits cross-task transfer, demonstrating a practical path toward consolidating large-scale discovery capabilities into a single language-steerable generative model. |
| title | A Unified Language Model for Large Scale Search, Recommendation, and Reasoning |
| topic | Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2603.17533 |