WRSWorld Research StateContribute
WORLD RESEARCH STATE

SPLADE: sparsity-regularized lexical expansion and distilled max-pooling variants

Contribution grade

Sparse neural information retrieval · Human-designed research method or resource

Reference score130,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Learned vocabulary-space query/document expansion with log-saturated weights and explicit sparsity regularization; v2 changed aggregation to max pooling, evaluated document-only expansion and added teacher-distilled hard-negative training.

Scientific scope

Learned vocabulary-space query/document expansion with log-saturated weights and explicit sparsity regularization; v2 changed aggregation to max pooling, evaluated document-only expansion and added teacher-distilled hard-negative training. FLOPS regularization and sparse expansion build on prior work, not invented wholesale here. Efficiency proxies, index storage and actual latency are distinct; offline preprocessing remains. Original preprints contain placeholder conference/DOI fields, not valid publication receipts. BEIR subset evaluations are not full benchmark-wide dominance.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. FLOPS regularization and sparse expansion build on prior work, not invented wholesale here. Efficiency proxies, index storage and actual latency are distinct; offline preprocessing remains. Original preprints contain placeholder conference/DOI fields, not valid publication receipts. BEIR subset evaluations are not full benchmark-wide dominance.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction