← 研踪 PaperTrailarXiv:2611.09680 · v1
SPARSEUP: Towards Better Recipes for Training Modern Learned Sparse Retrievers
首次提交 2026-11-07 00:27
版本更新 2026-11-07 00:27
雷达观测 2026-11-09 13:22
摘要
Learned sparse embeddings are a powerful alternative to dense retrieval, yet they have fallen behind single-vector and late interaction models. One reason for their limited adoption is that they are notoriously difficult to train, generally requiring very tight conditions to converge. We present a recipe for training Splade-like encoders on modern backbones and data that achieves a better efficiency-effectiveness trade-off. We discuss the role of logit distribution and its impact on the training stability and the final sparsity of the encoder. We also introduce a new per-position fan-out control - a hard cap on how many vocabulary dimensions each position may emit. We further apply a simple vocabulary merging strategy that removes redundant entries - such as case and word-boundary variants of the same word - which are common in cased BPE-based backbones. We apply our recipe to the LateOn backbone and training data, obtaining SPARSEUP, a sparse encoder that reaches 56.4 average nDCG@10 on BEIR-13, the highest we know of among public vocabulary-based sparse encoders below 150M parameters.
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-07 00:27 · 最新观测版本
SPARSEUP: Towards Better Recipes for Training Modern Learned Sparse Retrievers
Learned sparse embeddings are a powerful alternative to dense retrieval, yet they have fallen behind single-vector and late interaction models. One reason for their limited adoption is that they are notoriously difficult to train, generally requiring very tight conditions to converge. We present a recipe for training Splade-like encoders on modern backbones and data that achieves a better efficiency-effectiveness trade-off. We discuss the role of logit distribution and its impact on the training stability and the final sparsity of the encoder. We also introduce a new per-position fan-out control - a hard cap on how many vocabulary dimensions each position may emit. We further apply a simple vocabulary merging strategy that removes redundant entries - such as case and word-boundary variants of the same word - which are common in cased BPE-based backbones. We apply our recipe to the LateOn backbone and training data, obtaining SPARSEUP, a sparse encoder that reaches 56.4 average nDCG@10 on BEIR-13, the highest we know of among public vocabulary-based sparse encoders below 150M parameters.
查看此版本解读 →查看此版本原文 ↗