CoRoVA: Compressed Representations for Vector-Augmented Code Completion

TL;DR

CoRoVA introduces a method to compress context in retrieval-augmented code generation, significantly reducing latency and improving prediction quality with minimal additional computation.

Contribution

It presents a novel compression framework that maintains semantic richness and interpretability, enhancing code completion efficiency and effectiveness.

Findings

Achieves 20-38% reduction in time-to-first-token compared to uncompressed retrieval-augmented generation.

Requires only training a small projector module, adding negligible latency.

Improves generation quality while keeping prompt size minimal.

Abstract

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.