Salvato in:
Dettagli Bibliografici
Autori principali: Cherniuk, Daria, Sukhorukov, Nikita, Gusak, Danil, Sushko, Nikita, Sivtsov, Danil, Tutubalina, Elena, Frolov, Evgeny
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2510.19644
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917403998289920
author Cherniuk, Daria
Sukhorukov, Nikita
Gusak, Danil
Sushko, Nikita
Sivtsov, Danil
Tutubalina, Elena
Frolov, Evgeny
author_facet Cherniuk, Daria
Sukhorukov, Nikita
Gusak, Danil
Sushko, Nikita
Sivtsov, Danil
Tutubalina, Elena
Frolov, Evgeny
contents Retrieval-augmented generation has emerged as one of the most effective approaches for code completion enhancement, especially when repository-level context is important. However, adding this extra retrieved context significantly increases sequence length, raises prefill cost, and degrades time-to-first-token (TTFT), which slows down inference -- a critical limitation for interactive settings such as IDEs. In this work, we introduce CoRoVA, a framework that compresses context into compact, semantically rich representations that remain interpretable to code LLMs. This improves generation quality while reducing prompt augmentation to only a few compressed single-token vectors. Our approach requires training only a small projector module and introduces negligible additional latency, yet it significantly improves the prediction quality of code LLMs. Our experiments show that CoRoVA enables a 20-38\% reduction in TTFT on completion tasks compared to uncompressed RAG.
format Preprint
id arxiv_https___arxiv_org_abs_2510_19644
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CoRoVA: Compressed Representations for Vector-Augmented Code Completion
Cherniuk, Daria
Sukhorukov, Nikita
Gusak, Danil
Sushko, Nikita
Sivtsov, Danil
Tutubalina, Elena
Frolov, Evgeny
Computation and Language
Retrieval-augmented generation has emerged as one of the most effective approaches for code completion enhancement, especially when repository-level context is important. However, adding this extra retrieved context significantly increases sequence length, raises prefill cost, and degrades time-to-first-token (TTFT), which slows down inference -- a critical limitation for interactive settings such as IDEs. In this work, we introduce CoRoVA, a framework that compresses context into compact, semantically rich representations that remain interpretable to code LLMs. This improves generation quality while reducing prompt augmentation to only a few compressed single-token vectors. Our approach requires training only a small projector module and introduces negligible additional latency, yet it significantly improves the prediction quality of code LLMs. Our experiments show that CoRoVA enables a 20-38\% reduction in TTFT on completion tasks compared to uncompressed RAG.
title CoRoVA: Compressed Representations for Vector-Augmented Code Completion
topic Computation and Language
url https://arxiv.org/abs/2510.19644