AI Paper Atlas

Research map / Architectures and efficiency / Quantization

CrossQ: Task-Aligned Cross-Token Conditional Quantization for Late Interaction Retrieval

Rohit Kumar Salla, Manoj Saravanan, Ramya Manasa Amancherla

ICML 2026 · Architectures and efficiency › Quantization · approach: KV cache quantization · 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 document context h(D) to code selection (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.371 -> 0.386Improved
add listwise KL distillation (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.377 -> 0.386Improved
add ranking losses and context over reconstruction-only training (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.358 -> 0.386Improved
add context and listwise together (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.364 -> 0.386Improved
add context pathway versus parameter-matched no-context (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.372 -> 0.386Improved
use correct document context instead of shuffled context (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.370 -> 0.386Improved
use untied assignment keys (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.383 -> 0.386Improved
replace mean pooling with DeepSets context encoder (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.378 -> 0.386Improved
replace DeepSets with 1-layer Transformer context (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.386 -> 0.387No clear effect
initialize codebooks from OPQ instead of random Gaussian (ablation)CrossQ · MS MARCO dev (4 B/token)MRR@10: 0.386 -> 0.388No clear effect
add light fine-tuning of last two blocksCrossQ · MS MARCO dev (2 B/token)MRR@10: 0.369 -> 0.374Improved
apply fine-tuning to OPQ versus CrossQOPQ · MS MARCO dev (4 B/token)MRR@10: +0.002 vs +0.006Mixed

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