CS Paper Radar

VADER: Filtered Vector Search with Declarative Recall

Manos Chatzakis, Duo Lu, Helena Caminal, Yannis Chronis, Fatma Özcan, Yannis Papakonstantinou, Timofey Asyrkin, Sebastian Infante Murguia, Itai Rosenblatt, Themis Palpanas

cs.DB ·submitted 2026-09-29 17:22 UTC · v2
top venue

Abstract

Approximate filtered vector search (FVS), a core operation in many data management tasks that combine structured data with vector embeddings, exhibits increased complexity due to the characteristics of filtering predicates. Each predicate is defined by selectivity (i.e., the fraction of vectors that satisfy the predicate) and correlation (i.e., the relationship between the filter and the vector space), which can significantly affect search difficulty even for the same query vector. This poses a key challenge for users aiming to integrate vector search with structured data, as efficient execution often requires extensive manual tuning of algorithm parameters. In this paper, we present VADER, the first approach that eliminates hyperparameter tuning by introducing declarative recall for approximate filtered vector search. With declarative recall, users specify a desired recall target, and VADER executes FVS queries to meet this target without requiring manual configuration. VADER achieves this by employing a filter-aware recall predictor that generalizes across varying selectivities and correlations without explicit tuning, and by performing early termination once the predicted recall reaches the user-defined target. Through extensive experimental evaluation, we show that VADER achieves near-optimal early termination, while providing significant speedups of up to 53% faster and improved result quality of 28% compared to the best-performing baseline.

Comment
This paper appeared in VLDB 2027

Closest papers

Nearest neighbours by meaning across the whole index.

Analyzing and Reducing Search Quality Differences in Vector Similarity Search

cs.DB cs.OS ·2026-08-25 ·Sara Mahdizadeh Shahri, Martin Prammer, Jignesh M. Patel, Akshitha Sriraman
similarity 0.75
Abstract

Modern database services scalably search over large data collections via Approximate Nearest Neighbor Search, which improves search performance at the cost of search quality, measured by recall. In practice, a database operator seeks to achieve a target mean recall while maximizing throughput across search queries. We show that optimizing for mean recall masks significant differences in recall across queries even when target recall is met. As a result, numerous queries face (1) below-target recall, hurting user experience and revenue and (2) above-target recall, wasting computation to deliver unnecessarily high search quality. Thus, it is critical to detect and reduce recall differences across queries. We design RCheck, a light-weight run-time system that identifies low-recall queries and reduces recall differences while achieving high throughput. RCheck's key design principle is to dynamically, efficiently adapt search effort by increasing effort for queries below target recall and decreasing effort for those above it. RCheck tunes available search effort parameters, making it readily deployable. We evaluate RCheck using the widely-used production-style pgvector database. At the same throughput, RCheck improves mean recall by 11-93% and enables 8-47% more queries to meet target recall compared to the state-of-the-art globally-tuned configuration.

0.23

BOA: Beamwidth Online Adaptation for Filtered-ANNS on a GPU

cs.DC ·2026-09-14 ·Farhana Akter Tumpa, Rajiv Gupta
similarity 0.68
Abstract

Filtered approximate nearest neighbor search, i.e. returning the top-k vectors nearest to a query vector among those satisfying one or more attribute predicates, has become a fundamental operation in modern vector search systems. Graph-based solutions employ beam search to solve a batch of queries in parallel for high throughput and employ high fixed beamwidth of 100 or greater for ensuring high recall. We observe that, given a batch of queries, more than half of the queries across multiple data sets can be solved precisely with a beamwidth of just 50 or less. Therefore, existing systems based on fixed high beamwidth sacrifice throughput to achieve high recall by forcing every query to search as thoroughly as the hardest query in the batch even though majority of queries can be resolved by a shallow search. In this paper we present a filtered ANNS engine for a single GPU named BOA that uses online beamwidth adaptation to customize the search effort across queries within a batch under multi-attribute range filters. We address the recall throughput tradeoff with a multi-phase search: all queries are first evaluated under a narrow beam, and only those with uncertain results are progressively refined with wider beamwidths. This renders recall largely insensitive to the starting beamwidth, whereas prior methods must use a fixed high beamwidth for high recall. BOA+ overlaps execution of phases to further enhance throughput. Our experiments show that, for 10,000 queries, online adaptation achieves 94.05% to 99.96% recall with average beamwidth ranging from 22 to 77, while a non-adaptive approach requires a fixed beamwidth of 500 to achieve similar or lower recall. Consequently, adaptivity increases throughput by 7x to 12.5x

0.07

KVFetch: Temporal Prefetching for the Missing Half of KV Cache Compression

cs.LG cs.AI ·2026-09-23 ·Linfeng Dong
similarity 0.65
Abstract · 21 pages, 7 figures. Submitted to ICLR 2027

As context windows scale to tens or hundreds of thousands of tokens, KV cache compression has become essential for efficient LLM inference. Existing methods fall into three families: score-based eviction, summary compensation, and offload-and-recall. Yet all three decide what to keep or recall by content relevance to the current query. We show this shared design is structurally incomplete. A cache supports two access modes: associative lookup by content and sequential traversal by position; current compressors implement only the first. The gap matters in practice: retrieval-augmented generation, code completion, and structured-data extraction all require the model to reproduce identifiers, field values, or code tokens verbatim from the context. Under compression, content-based eviction retains the head of such a sequence but discards its continuation, causing verbatim copying to break irreversibly midway, a failure we call sequential forgetting. This failure resists better scoring, larger budgets, summary compensation, and dynamic re-scoring; it is the dominant source of remaining quality loss under compression. We propose KVFetch, a training-free, drop-in framework that opens a temporal recall channel for any score-based compressor. It demotes evicted candidates to a quantized cold tier, detects active copying through a monotone read pointer, and prefetches positional successors into fixed-size hot-tier slots without increasing attention cost. On RULER-16K under an iso-budget control, KVFetch recovers verbatim copying from 0.8 to 78.4 and raises the 13-task average by +8.4, with gains concentrating on tasks that require sequential access. On LongBench, where no task requires sequential access, the channel remains dormant and imposes no cost.

0.18

Fidelity Synopses: Pre-Execution Plan Selection for Approximate Vector Search

cs.DB cs.IR ·2026-10-30 ·Adithya Ramanatha, Sanskar S. Singh, Maya Ramanath
benchmark similarity 0.65
Abstract

Vector search has become a database operator with many physical implementations: different index families, build-time and search-time parameters, and compressed or disk-resident variants. Each trades cost against fidelity, the fraction of a query's exact top-k nearest neighbors that a configuration returns. Deployments nevertheless choose one configuration offline and use it for every query, so they are sized for their hardest queries and overpay for the rest. On a 5.8M-item LAION benchmark with four embedding models, running each query on the cheapest configuration that suffices for it would roughly halve the cost in a typical setting. Exploiting this requires predicting each configuration's fidelity before running it, which is hard because fidelity is defined by the very exact search that approximate indexes avoid. Relational optimizers solve the analogous problem for cardinality with offline synopses; we propose one for vector search. A fidelity synopsis divides the embedding space into cells around a set of landmarks and records, for every cell and configuration, how often that configuration met a fidelity target on a sample of queries. At query time, one nearest-landmark lookup yields a calibrated estimate for every configuration, and an optimizer chooses configurations so that the workload misses the target no more often than its requirement allows. We test the two premises the synopsis relies on, that fidelity varies smoothly across the embedding space and that its loss is specific to each configuration. Held to the same requirement a static deployment is built to meet, synopsis-guided selection lowers cost in all 16 combinations of model and query direction, to a median of 0.72x the static deployment, with a batched lookup of 21-32 microseconds per query. The saving persists against the best static choice among the same indexes, and within a single index at no extra memory.

0.09

Signals and Queries: Token-Efficient Agentic Context Discovery

cs.CL ·2026-10-21 ·Yam Marcovitz
similarity 0.64
Abstract

User-facing LLM agents must identify which behavioral policies apply at each turn, but evaluating every policy with an LLM is costly. We introduce Signals and Queries (S&Q), a lightweight first-stage relevance filter developed for Parlant 4.0 that removes at least 30% of candidate policies before downstream evaluation while preserving high recall. For each policy, an LLM generates activating examples and hard negatives offline. An independent logistic classifier trained on their embeddings then predicts relevance using embeddings and vector dot products, with no per-turn LLM calls. This shifts expensive relevance evaluation to a smaller candidate set, while allowing each policy's classifier to be updated independently. Evaluated against 9,144 LLM-judge relevance labels covering 254 $τ^2$-bench conversation prefixes and 36 airline policies, S&Q achieves higher precision than top-K and cosine-threshold retrieval at every evaluated recall level, with consistent results across two embedding providers. These findings support inexpensive, per-policy classification as a practical first stage for reducing downstream evaluation workload in customer-facing agents.

0.23

Less can be More: Relieving RAG Bottlenecks via Evidence Frontloading and Pressure-Adaptive Budgeting

cs.CL cs.IR ·2026-08-25 ·Weibin Cai, Reza Zafarani
similarity 0.63
Abstract

Existing methods for improving Retrieval-Augmented Generation (RAG) efficiency mainly optimize downstream LLM generation, such as context compression or serving optimization. However, RAG is an end-to-end system, and its bottleneck can shift between upstream reranking and downstream generation under different serving loads and reranking budgets.In this paper, we first empirically characterize this shifting-bottleneck behavior and show that upstream reranking can become the dominant bottleneck under high query rates or large reranking budgets. Reducing the reranking budget can relieve this bottleneck, but it may also drop supporting evidence and degrade recall. To address this problem, we propose \textbf{\textsf{PACE}} (\textbf{P}rioritized \textbf{A}daptive \textbf{C}overage of \textbf{E}vidence), a training-free framework that combines \textit{evidence frontloading} with \textit{pressure-adaptive budgeting}. \textsf{PACE} first reorders candidates by marginal evidence coverage, prioritizing documents that are query-relevant, complementary, and useful for forming multi-hop evidence chains. We show that this objective is monotone submodular, giving greedy selection a $(1-1/e)$ approximation guarantee. \textsf{PACE} then dynamically adjusts the reranking budget according to the relative pressure of the reranker and the LLM. Experiments on three multi-hop QA datasets and online serving simulations show that \textsf{PACE} improves evidence recall, reduces p95 latency under ranking-heavy workloads. More importantly, the two components together reveal that \textit{less can be more}: an evidence-dense top-ranked candidates enable higher final recall with fewer reranked documents.

0.36

Calibrated Early Termination with Bounded Optimality Gaps for Constrained Vector Search

cs.DB cs.IRcs.LG ·2026-11-12 ·Xinyu Zhang, Jingze Li, Baihua Zheng, Weiwei Sun
theorycode/data similarity 0.63
Abstract · 14 pages, 5 figures, 3 tables. Code available at https://github.com/zhangshea/CaST

Vector search over learned embeddings is the retrieval layer of recommendation and retrieval-augmented generation, and many applications cap how many answer items one category may contribute. Graph indexes answer a query by best-first traversal, and where it stops determines its cost and whether its answer is exact. We study early termination for such constrained $k$-nearest-neighbor ($k$-NN) queries, with plain $k$-NN as a special case, and seek the cheapest stopping rule whose answer is exact with probability at least $1-\varepsilon$. Learned early-termination methods instead target average recall, which does not guarantee exact answers. Multiple testing can calibrate any parameterized stopping rule to our target, but it cannot tell how much a better rule with the same risk and observations would save. Answering this needs a lower bound on the optimum of a rule family too large to enumerate, valid for the query population. Our key idea is to declare the observations a rule may use and to consider every rule monotone in them. Risk-constrained stopping then becomes a linear program over an order polytope, whose Lagrangian minimizers form a nested chain of staircase rules computed exactly by dynamic programming. Fixed-sequence binomial tests along this chain calibrate our rule, CaST, whose failure probability is at most $\varepsilon$ with confidence $1-δ$. Lagrangian weak duality bounds the optimum of the declared class from below at all risk levels at once, without a union bound over the rules. On two datasets of one million items each, CaST replayed on logged searches costs at most 1.132 times the class optimum for risk levels from 0.03 to 0.10, at 95% joint confidence per setting. Larger savings thus need a rule outside the class. At a common risk target on held-out queries, CaST is faster than every other calibrated rule we time, which take 4.8% to 45.5% more mean time.

0.28

AiSAQ-in-Filtering for DRAM-free Filtered Information Retrieval

cs.IR ·2026-10-13 ·Kento Tatsuno, Daisuke Miyashita, Kazunari Sumiyoshi, Hiroaki Yamanaka, Kosuke Nakata +1 more
similarity 0.63
Abstract · 5 pages, 7 figures

Filtering search is one of the crucial applications of approximate nearest neighbor search (ANNS). In filtering search, label information is used in conjunction with unstructured vector data. ANNS provides efficient similarity search in large-scale datasets, however, the memory footprint of storing both vector data and their associated label becomes a significant cost factor as dataset sizes grow. To address this problem of increasing DRAM usage, we propose AiSAQ-in-Filtering, with a novel quantization method called LiPQ (Label-inclusive Product Quantization). LiPQ combines unstructured data and structured labels into a single representation, leading no growth of encoded data. Our proposal increased recall@1 of single-label searches by up to 55% and fewer disk I/Os compared with existing methods. This also retains the main advantages of AiSAQ, which can reduce DRAM footprint by up to 1/78 and index switching time comparing with existing methods. This proposal can also be applied to sparse-dense hybrid search beyond label filtering such as semantic and keyword search of documents.

0.17