cs.LGMay 22, 2026

Is Dimensionality a Barrier for Retrieval Models?

Authors: Kiril BangachevGuy BreslerJonathan KoganYury Polyanskiy

Abstract

Why does the low dimensionality of representations, typically d1000d\approx 1000, not prevent modern embedding-based retrieval models from scaling to billions, or even trillions, of data points? To answer this question, we study maximal-margin embeddings in the following retrieval model, classically studied in communication complexity [PS86] and more recently in embedding-based retrieval [WBNL26]. Let A{0,1}N×nA\in \{0,1\}^{N\times n} be a matrix indicating whether each of NN queries is relevant to each of nn documents. We are interested in the largest margin m>0,m>0, denoted by mrd(d,A),\mathsf{m}^{\mathsf{rd}}(d, A), for which there exist unit norm embeddings of the queries and documents {Uj}j=1N,{Vi}i=1n\{U_j\}_{j = 1}^N, \{V_i\}_{i = 1}^n with the following property. Uj,Vim\langle U_j, V_i\rangle \ge m whenever Aji=1A_{ji} = 1 and Uj,Vim\langle U_j, V_i\rangle \le -m otherwise. A large margin is a key proxy for representation quality: it controls both robustness to perturbations and compositional generalization across queries. Our main theorem establishes that the best possible margin without a restriction on the dimension, mrd(+,A),\mathsf{m}^{\mathsf{rd}}(+\infty, A), can be nearly achieved in dimension d=O(mrd(+,A)2logn)d = O(\mathsf{m}^{\mathsf{rd}}(+\infty, A)^{-2}\log n) which improves a theorem of [BDES02]. Together with a matching lower bound in Theorem 1.5, we conclude that when A{0,1}(nk)×nA\in \{0,1\}^{\binom{n}{k}\times n} is the matrix containing all possible kk-sparse rows once, dimension d=O(klog(n/k))d = O(k\log (n/k)) is necessary and sufficient for the maximal possible margin mrd(+,A)=Θ(k1/2)\mathsf{m}^{\mathsf{rd}}(+\infty, A) = Θ(k^{-1/2}) in this setting. This fully resolves the setup of [WBNL26]. We also give several constructions for large margins when d=o(klog(n/k)).d = o(k\log (n/k)). Finally, we empirically test the InfoNCE and sigmoid losses for producing large margin embeddings and demonstrate a clear advantage of the sigmoid loss.

Explore similar work

Jun 10, 2026cs.IR

What Limits Does Quantization Place on Dense Top-k Retrieval? A Theoretical Study

We establish conditions for embedding a corpus of NN documents as dd-dimensional vectors such that every kk-subset S[N]S \subseteq [N] is realizable as a result of top-kk retrieval by some query vector. Recent work shows that d=O(k)d = O(k) suffices for such embeddings to exist in Rd\mathbb{R}^d, independently of NN. We theoretically prove that this corpus-independent bound is specific to infinite precision. With BB bits per coordinate, perfect top-kk retrieval requires Bd=Ω(klnN)Bd = Ω(k \ln N); thus, at any fixed precision, the dimension must grow at least logarithmically with NN. Specializing to a 2\ell_2-normalized BB-bit uniform scalar quantization model, we also identify a threshold on the precision B=O(lnlnN)B^{*} = O(\ln \ln N) below which no dimension suffices, together with two further regimes that bound the feasible (B,d)(B, d) pairs. Our result implies that in practical vector databases and dense retrieval systems where quantization is standard, the embedding dimension and possibly the precision must grow with the corpus size.
Koki Okajima, Tsukasa Yoshida
Jun 9, 2026cs.IR

The Voronoi Bottleneck: Capacity-Aware Dense Retrieval for Product Search

Dense embedding retrieval compresses all relevance information into a single inner product, imposing a fundamental geometric limit -- the Voronoi Bottleneck -- on the number of query-document relevance patterns expressible at fixed embedding dimension (d). We make three contributions. (1) Unified capacity theory. We prove that Voronoi complexity and sign-rank are equivalent for top-1 retrieval, yielding tight dimension bounds and a computable diagnostic, the Capacity Utilization Score (CUS), that predicts per-query retrieval failure with AUC (> 0.8) without relevance labels. (2) Diagnosis. CUS identifies two capacity regimes -- moderate ((δ\gtrsim 1)), where density-aware training yields measurable gains, and vacuous ((δ\ll 1)), where it does not -- giving practitioners an a priori check before investing in retraining. (3) DART training. We introduce AT-DW-InfoNCE, an Adaptive-Temperature Density-Weighted contrastive objective with formally derived optimal weighting (α^* = 2.0). On a 100K-query synthetic product-search corpus with controlled relevance structure, DART improves +1.9 Recall@100 over a same-data InfoNCE baseline ((84.9 \pm 0.0) vs. (83.0 \pm 0.3); 8 seeds, (p < 0.001)), outperforming focal loss and temperature-schedule alternatives. DART requires zero inference-time overhead -- it is a drop-in training objective that improves any dual-encoder system.
Charith Chandra Sai Balne, Rithwik Maramraju, Siddharth Pratap Singh +4
Jul 3, 2026cs.IR

Relevance-Based Embeddings: Lightweight Candidate Retrieval via Heavy-Ranker Calls

In many machine learning applications, the most relevant items for a query should be efficiently retrieved. The relevance function is usually an expensive similarity model, making the exhaustive search infeasible. A typical solution is to train another model that separately embeds queries and items to a vector space, where similarity is defined via the dot product or cosine similarity. This allows one to search the relevant items through fast approximate nearest neighbor search at the cost of some reduction in quality. To compensate for this reduction, the found items (candidates) are re-ranked by the expensive ranking model. In this paper, we investigate an alternative approach to candidate selection that utilizes the scores of the expensive model to improve the representations of queries and items. The idea is to describe each query (item) by its relevance to a set of support items (queries) and use these new representations to obtain query (item) embeddings. We theoretically prove that such embeddings are powerful enough to approximate any complex similarity model (under mild conditions). We also investigate the choice of support items, which is a crucial ingredient of the proposed approach. The experiments on diverse academic and production datasets illustrate the power of our method.
Kirill Shevkunov, Andrey Ploskonosov, Liudmila Prokhorenkova