collaborators

10 papers

cs.IR2026

QASP: Query-Adaptive Robust Vector Search Policy

Hakan Ferhatosmanoglu, Kushal Kumar, Tal Wagner +1

A fundamental challenge of vector search is achieving consistently high recall while minimizing computational costs. Fixed search parameters cause significant performance variance…

cs.LG2026

Frustratingly Simple Black-Box Adaptation of Language Models via Logit Bias

Ofek I. Cohen, Lior Shani, Aviv Rosenberg +3

Many organizations aim to adapt language models for internal use, both to improve performance on domain-specific tasks and to address privacy concerns around sensitive data. Howeve…

cs.LG2026

Positional LSH: Binary Block Matrix Approximation for Attention with Linear Biases

Daniel Wolfson, Tal Wagner

Positional encoding in transformers is commonly implemented through positional embeddings, attention masks, or bias terms, but formal connections between these mechanisms remain li…

cs.DS2026

New Bounds for Kernel Sums via Fast Spherical Embeddings

Tal Wagner

We study query time bounds for the fundamental problem of estimating the kernel mean of a query in a finite dataset $X\subset\mathbb{R…

cs.DS2025

Quantization for Vector Search under Streaming Updates

Ishaq Aden-Ali, Hakan Ferhatosmanoglu, Alexander Greaves-Tunnell +2

Large-scale vector databases for approximate nearest neighbor (ANN) search typically store a quantized dataset in main memory for fast access, and full precision data on remote dis…

cs.DS2025

Graph-based Nearest Neighbors with Dynamic Updates via Random Walks

Nina Mishra, Yonatan Naamad, Tal Wagner +1

Approximate nearest neighbor search (ANN) is a common way to retrieve relevant search results, especially now in the context of large language models and retrieval augmented genera…