Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2607.08107

BACH: A Bayesian Admixture of Contrastive Heads for Multi-Interest Two-Tower Retrieval

Quoc Phong Nguyen·Paul Albert·Long Vuong·Vuong Le·Julien Monteil
View original source ↗Read source document ↗Export metadata ↓

Abstract

Two-tower retrievers compress each user into a single embedding, limiting their ability to serve diverse interests. Multi-interest models give each user several heads scored by a maximum inner product, but their hard-routing training under-utilizes heads (routing collapse) and gives no per-user estimate of how much each interest matters for serving. We present \textbf{BACH} (\emph{Bayesian Admixture of Contrastive Heads}), which casts multi-interest two-tower retrieval as a per-user mixture over the heads, fit by variational inference. The soft mixture trains every head (mitigating collapse), produces a per-user weighting of the interests that is reused at serving, and admits a shared global-codebook variant with precomputable retrieval. On three large-scale benchmarks, MovieLens-20M, Taobao, and Netflix, BACH improves top-of-ranking retrieval over hard-routing multi-interest and single-vector baselines at every head count; we further find that scoring every candidate by its best head, consistent with serving, outperforms the usual target-routed training, and that BACH improves further still.

Explore related subjects

cs.IRcs.LG

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Quoc Phong Nguyen, Paul Albert, Long Vuong, Vuong Le, Julien Monteil. 2026-07-09. BACH: A Bayesian Admixture of Contrastive Heads for Multi-Interest Two-Tower Retrieval. https://arxiv.org/abs/2607.08107

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Retrieval-in-the-Chain: Bootstrapping Large Language Models for Generative Retrieval

Generative retrieval (GR) is an emerging paradigm that leverages large language models (LLMs) to autoregressively generate document identifiers (docids) relevant to a given query. Prior works have focused on leveraging the generative capabilities of LLMs to improve GR, while overlooking that their reasoning capabilities could likewise help. This raises a key question: Can explicit reasoning benefit GR? To investigate, we first conduct a preliminary study where an LLM is prompted to generate free-form chain-of-thought (CoT) reasoning before performing constrained docid decoding. Although this method outperforms standard GR, the generated reasoning tends to be verbose and poorly aligned with the docid space. These limitations motivate the development of a reasoning mechanism better tailored to GR. Therefore, we propose Reason-for-Retrieval (R4R), a reasoning-augmented framework for GR that converts free-form CoT reasoning into a compact, structured format, and iteratively refines the reasoning during the retrieval process. R4R augments an existing GR method by leveraging a reasoning-capable LLM that has been instruction-tuned for GR. At inference time, R4R first uses the LLM to generate an initial structured reasoning; then the same LLM alternates between (i) constrained decoding with the chosen GR method to produce candidate docids and (ii) updating the reasoning based on retrieval results to improve the next round. R4R does not require additional models or training, and instead a single LLM serves as both the reasoning generator and the retriever. Extensive experiments on Natural Questions, MS MARCO, and a real-world item-search benchmark validate the effectiveness of R4R.

cs.IR↗
arXiv2026

Understanding Mobile App Recommendation Dynamics in General-Purpose LLMs: An Empirical Study

Large Language Models (LLMs) are increasingly used to recommend mobile applications through natural language prompts, offering a flexible alternative to keyword-based app store search. Yet, the reasoning behind these recommendations remains opaque, raising questions about their consistency, explainability, and alignment with traditional App Store Optimization (ASO) metrics. In this paper, we present an empirical observational study of how general-purpose LLMs generate, justify, and rank mobile app recommendations across proprietary and open-source models, as well as knowledge-only settings with web search evaluated as a controlled ablation on the proprietary cohort. Our contributions are: (i) a taxonomy of 16 generalizable ranking criteria elicited from LLM outputs; (ii) a systematic evaluation framework to analyse recommendation consistency and the effect of explicit ranking instructions on cross-model convergence; and (iii) a replication package to support reproducibility and future research on LLM-based recommendation systems. Our findings reveal that LLMs report a broad yet fragmented set of ranking criteria, only partially aligned with standard ASO metrics. Proprietary models produce substantially more stable recommendations than locally deployed open-source models, and consistency varies substantially across app domains. Furthermore, enabling web search does not materially change the ranking criteria reported by LLMs. Contrary to our hypothesis, conditioning on explicit ranking criteria steers recommendations away from the blind baseline but reduces rather than increasing cross-model convergence. Our results aim to support end-users, app developers, and recommender-systems researchers in navigating the emerging landscape of conversational app discovery.

cs.IR↗
arXiv2026

Personalized and Trust-Aware Health Recommendation Policies for a Construction Workplace

Construction workers face workplace risks such as fatigue, heat stress, and other physically demanding conditions that can negatively affect their health and safety. Although monitoring these risks is important, timely and personalized health interventions are also needed to help prevent negative impacts on workers' well-being and productivity. To this end, in this paper, we propose a model to capture the interactions between a trust-aware health recommender system and workers who differ in health and trust sensitivity. Specifically, in our proposed dynamic model, worker health evolves over time, worker trust is affected by both health and recommendation dynamics, and trust in turn affects compliance with future recommendations. Given this model, we characterize the recommender policy, including a health-based recommendation triggering threshold and the recommendation frequency. We do so using both model-based short-horizon control and model-free reinforcement learning. We then investigate how recommendation frequencies are adjusted for different workers to balance their health, productivity, and trust. Our findings provide insight into the design of personalized health recommendation policies in construction workplaces and beyond.

cs.IR↗