Alexandre Salle, Chenglei Niu, Suchismit Mahapatra, Xiaoxiao Chen, Suvash Sedhain, Yaqi Wang, Shervin Shahryari, Saurabh Agrawal, Qiang Chen, Michael Tamir · · 2026
DOI: 10.1145/3773078.3831899
Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).
Personalized discovery systems often train separate models for item ranking, carousel ranking, and search, even though these tasks expose complementary signals from the same viewer journey: watches shape carousel and item ranking, search queries reveal intent even when they do not lead to a catalog match, and watch history helps interpret search as rewatching, continuation, or new discovery. We introduce the user story, a serialized representation that turns a user’s cross-surface history (attributes, sessions, watch events with surface and carousel context, and search events) into a single token sequence. By interleaving pretrained language tokens with domain-specific event tokens, user stories let heterogeneous recommendation and search tasks be expressed as prompted next-token prediction over a shared grammar. TubiFM is one instantiation of this approach: a Llama 3.2 1B-based model trained on user stories and prompted to rank items, carousels, or search results without task-specific architectures. In offline evaluation, this single model outperforms specialist baselines across item, carousel, and search ranking. In online A/B tests, TubiFM significantly improves search total viewing time (TVT) by +3.9% and carousel TVT by +0.30%. Item ranking is statistically neutral on TVT (+0.14%), matching a mature production stack while replacing it with a single model. Across all three tasks, TubiFM serves on L40S GPUs and reduces p99 ranking latency from approximately 500ms to 200ms. These results show that shared user stories can improve discovery while simplifying ranking systems.
No comments yet — start the discussion below.