AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

No More K-means: Single-Stage Sparse Coding for Efficient Multi-Vector Retrieval

Lixuan Guo, Yifei Wang, Tiansheng Wen, Aosong Feng, Stefanie Jegelka, Chenyu You

ICML 2026 · Architectures and efficiency › RNNs and normalization · approach: Learned embeddings for retrieval · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add SSR++ coarse-to-fine pruning (ablation)SSR · MS MARCO passage subsetretrieval latency (ms): 38.6 -> 17.5Improved
add SSR++ coarse-to-fine pruning (ablation)SSR · MS MARCO passage subsetcandidates hit: 54,278 -> 3,196Improved
add SSR++ coarse-to-fine pruning (ablation)SSR · MS MARCO passage subsetnDCG@10: 45.3 -> 45.2No clear effect
use SSR sparse late interaction instead of Splade-v3SSR-CLS · BEIR 13 datasetsnDCG@10: 51.2 -> 53.4Improved
use SSR-tok instead of ColBERTv2 for retrievalSSR-tok · MS MARCO passageretrieval latency (ms): 56.9 -> 17.5Improved
apply SSR to frozen Llama-embed-nemotron-8bSSR-CLS · BEIR 13 datasets (avg)nDCG@10: 65.8 -> 67.1Improved
use SSR instead of a linear projector on frozen backbone (ablation)Llama-8B · BEIR 13 datasets (avg)nDCG@10: 66.1 -> 67.1Improved
add a linear projector to frozen e5-mistral-7b (ablation)e5-mistral-7b · BEIR 13 datasets (avg)nDCG@10: 58.8 -> 60.2Improved
use SSR instead of ColBERTv2 on LoTTE search queriesSSR · LoTTE Search (avg)Success@5: 71.3 -> 73.8Improved
use SSR instead of ColBERTv2 on LoTTE forum queriesSSR · LoTTE Forum (avg)Success@5: 65.1 -> 67.3Improved
increase hidden dimension h from 2^12 to 2^16 (ablation)SSR · BEIRnDCG@10Mixed
reduce sparsity K below 16 (ablation)SSR · BEIRnDCG@10Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy