Metraipolis › archive › dc_a3e6a2e58eee

PALATE: Pretrained Autoregressive LATent Embeddings — arXiv:2610.39090

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.39090
License (as archived)
CC0-1.0
Archived
2026-11-01T02:27:05.891Z
SHA-256 of the archived text
97879476c3215baef6f0622fec5a70674e9ff17d16d9031d5216e1d9aa3516cf
Stable URL of this page
https://metraipolis.com/documento/dc_a3e6a2e58eee

How to cite

PALATE: Pretrained Autoregressive LATent Embeddings — arXiv:2610.39090 — arXiv (en), https://arxiv.org/abs/2610.39090 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a3e6a2e58eee (Metraipolis archive dc_a3e6a2e58eee, sha256 97879476c3215baef6f0622fec5a70674e9ff17d16d9031d5216e1d9aa3516cf)
BibTeX
@misc{dca3e6a2e58eee,
  title = {PALATE: Pretrained Autoregressive LATent Embeddings — arXiv:2610.39090},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.39090},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a3e6a2e58eee; sha256 97879476c3215baef6f0622fec5a70674e9ff17d16d9031d5216e1d9aa3516cf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39090. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Local-commerce and food-delivery marketplaces run large ads businesses: across hundreds of thousands of stores, merchants pay to be surfaced on browse and search surfaces, and a ranker decides which ads each consumer sees. Ranking well demands strong per-consumer personalization, yet a consumer's raw interaction history is sparse and high-cardinality. Our production ranker had personalized only through hand-engineered signals: aggregate engagement counts, categorical embeddings, and order-agnostic pooled history, with no learned per-consumer sequence representation. We learn one by training a GPT-style generative model of the consumer's engagement sequence to predict their next few engagements, and read its last-token hidden state as the consumer embedding. That single frozen vector becomes one shared personalization input for every ad surface: a daily job refreshes it, and each ranker consumes it as an ordinary dense feature. Offline it is the largest single-feature gain in our ranking-AUC ladder, a relative Eval-AUC gain of +0.59% on a browse-driven feed and +0.57% on a query-driven search surface. Online, it shipped as one component of bundled ranker retrains on both surfaces: Discovery (revenue +0.79%, conversion rate +1.17%) and Search (revenue +0.94%, conversion rate +1.50%). Three design choices adapt the generative objective to a repurchase-heavy marketplace, and are the core of our contribution: an exploration-weighted loss that up-weights businesses new to the consumer over ones already engaged; a two-stage curriculum: a click pretrain that keeps the representation broad and exploratory, then a conversion fine-tune that sharpens purchase intent; and a designed negative set that pairs a fixed popular-business pool and businesses shown but not ordered with random and in-batch negatives.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Information Retrieval (cs.IR) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.39090 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-01T02:12:44.286Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).