1.  [](/)
2.  [Research](/research)
3.  Cluster with Auctions: The FAISS Team Teaches Vector Indexes to Learn From Queries

[Back to Research](/research)

Meta AI (FAISS)[arXiv:2607.13728](https://mxp.co/e/cwa)

# Cluster with Auctions: The FAISS Team Teaches Vector Indexes to Learn From Queries

Most IVF vector indexes assign queries with the same function that partitioned the data. Cluster with Auctions jointly learns a balanced partition and a neural probing function, reassigning vectors by a large-scale auction while the probe learns the live query distribution.

## Where this fits in Mixpeek

Mixpeek's vector indexes use IVF/ANN partitioning to search millions of embeddings in milliseconds. How the index partitions data and picks which partitions to probe is exactly what determines recall-versus-latency. Query-aware partitioning like this is the research frontier behind tuning Mixpeek's index configs: balanced clusters plus a probe that matches real query traffic get more recall from the same nprobe budget.

## About this research

IVF-style vector search partitions data into clusters and probes a few per query, but almost every system assigns queries with the same function that partitioned the data. Cluster with Auctions (Meta AI, the FAISS lineage) jointly learns a balanced partition and a neural probing function: vectors are reassigned by a large-scale auction while the probe learns the live query distribution. Paper: arxiv.org/abs/2607.13728

vector-searchannivfindexingfaiss

## Frequently asked questions

Why does the query distribution matter for a vector index?+

## Go deeper

*   [Diagram: Vector index structures →](/diagrams/vector-index-structures)
*   [Diagram: Storage vs. index →](/diagrams/storage-vs-index)

Put the research to work

Mixpeek turns video, images, audio, and documents in your object storage into searchable, timestamped results through one API: the retrieval stack these papers describe.

[Search your own data](https://studio.mixpeek.com/namespaces/create?mode=managed&utm_source=homepage&utm_medium=research_detail&utm_campaign=mvs_onramp)

## Related research

[

![Google's MUVERA: The One-Vector Trick Behind Fast Multi-Vector Search](https://cdn.mixpeek.com/videos/thumbs/78-muvera.jpg)

Google Research

Google's MUVERA: The One-Vector Trick Behind Fast Multi-Vector Search



](/research/muvera)[

![Learned Indexes Explained: When a Model Replaces the B-Tree](https://cdn.mixpeek.com/videos/thumbs/46-learned-indexes.jpg)

Google

Learned Indexes Explained: When a Model Replaces the B-Tree



](/research/learned-indexes)[

![How Meta Cut Its Ads Model Serving Cost From 24.6x to 3.9x (HILL, Explained)](https://cdn.mixpeek.com/videos/thumbs/63-meta-hill.jpg)

Meta

How Meta Cut Its Ads Model Serving Cost From 24.6x to 3.9x (HILL, Explained)



](/research/meta-hill-ads)