6 papers · 1 filter
Simple Linear Corrections for Compression in Late Interaction
Mihir Agarwal, Sabyasachi Basu, Viraj Agrawal
Late-interaction (multi-vector) models represent documents using a set of token-level vectors, providing a middle ground between cross-encoders and dense encoders in terms of quali…
EigenLI: Spectral Approximations to Late Interaction
Archish S, Sabyasachi Basu, Ankit Garg +2
Late-interaction models such as ColBERT achieve strong effectiveness by representing each document with many token-level vectors, but this expressivity leads to large indexing cost…
Finding Many Overlapping Dense Subgraphs Using Triadic Cohorts
Sabyasachi Basu, C. Seshadhri
Graphs are a standard representation for data in the social sciences, cybersecurity, computer infrastructure, bioinformatics, and more. Typical real-world graphs are sparse, meanin…
Retrieval Needs Multivectors: An Exponential Separation
Mihir Agarwal, Viraj Agrawal, Sabyasachi Basu +2
Recent works have highlighted the expressive limitations of embedding based retrieval models through both theoretical analyses and challenging benchmarks such as LIMIT. While multi…
Quantization Beyond Uniform Bit Allocation
K. S. Sreeramji, Sabyasachi Basu, Ravishankar Krishnaswamy +2
Quantization is a fundamental technique to handle the growing sizes of embeddings generated by modern models. Existing quantization schemes are largely embedding agnostic and alloc…
Query-Limited Community Recovery in Stochastic Block Models
Sabyasachi Basu, Manuj Mukherjee, Lutz Oettershagen +1
We study exact community recovery in the two-community stochastic block model on vertices under limited and noisy access to network data. The learner may query a noisy neighbor…