# Multi-line AI-assisted Code Authoring

arXiv: https://arxiv.org/abs/2402.04141 · HTML: https://mlsanity.com/paper/2402.04141/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 24 references that are in the map.
- First posted on arXiv 6 Feb 2024.

## Builds on (most cited first, 7 of 7)

- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [InCoder: A Generative Model for Code Infilling and Synthesis](https://mlsanity.com/paper/2204.05999/) (2022)
- [Grounded Copilot: How Programmers Interact with Code-Generating Models](https://mlsanity.com/paper/2206.15000/) (2022)
- [Code Prediction by Feeding Trees to Transformers](https://mlsanity.com/paper/2003.13848/) (2020)
- [Don't Complete It! Preventing Unhelpful Code Completion for Productive and Sustainable Neural Code Completion Systems](https://mlsanity.com/paper/2209.05948/) (2022)
- [Improving Code Autocompletion with Transfer Learning](https://mlsanity.com/paper/2105.05991/) (2021)

## Built on by (most cited first, 13 of 13)

- [Can LLMs Replace Manual Annotation of Software Engineering Artifacts?](https://mlsanity.com/paper/2408.05534/) (2024)
- [IRCoder: Intermediate Representations Make Language Models Robust Multilingual Code Generators](https://mlsanity.com/paper/2403.03894/) (2024)
- [textttDroid: A Resource Suite for AI-Generated Code Detection](https://mlsanity.com/paper/2507.10583/) (2025)
- [Rethinking Software Engineering in the Foundation Model Era: A Curated Catalogue of Challenges in the Development of Trustworthy FMware](https://mlsanity.com/paper/2402.15943/) (2024)
- [Agentic Program Repair from Test Failures at Scale: A Neuro-symbolic approach with static analysis and test execution feedback](https://mlsanity.com/paper/2507.18755/) (2025)
- [Human-AI Experience in Integrated Development Environments: A Systematic Literature Review](https://mlsanity.com/paper/2503.06195/) (2025)
- [ObscuraCoder: Powering Efficient Code LM Pre-Training Via Obfuscation Grounding](https://mlsanity.com/paper/2504.00019/) (2025)
- [Leveraging AI for Productive and Trustworthy HPC Software: Challenges and Research Directions](https://mlsanity.com/paper/2505.08135/) (2025)
- [Mellum: Production-Grade in-IDE Contextual Code Completion with Multi-File Project Understanding](https://mlsanity.com/paper/2510.05788/) (2025)
- [AI-Assisted Fixes to Code Review Comments at Scale](https://mlsanity.com/paper/2507.13499/) (2025)
- [Smart Paste: Automatically Fixing Copy/Paste for Google Developers](https://mlsanity.com/paper/2510.03843/) (2025)
- [The Perfection Paradox: From Architect to Curator in AI-Assisted API Design](https://mlsanity.com/paper/2603.12475/) (2026)
- [StreamDecisionBench: Evaluating Decisions in Force on Evolving Language Streams](https://mlsanity.com/paper/2609.38612/) (2026)

## Abstract (arXiv)

CodeCompose is an AI-assisted code authoring tool powered by large language models (LLMs) that provides inline suggestions to 10's of thousands of developers at Meta. In this paper, we present how we scaled the product from displaying single-line suggestions to multi-line suggestions. This evolution required us to overcome several unique challenges in improving the usability of these suggestions for developers. First, we discuss how multi-line suggestions can have a 'jarring' effect, as the LLM's suggestions constantly move around the developer's existing code, which would otherwise result in decreased productivity and satisfaction. Second, multi-line suggestions take significantly longer to generate; hence we present several innovative investments we made to reduce the perceived latency for users. These model-hosting optimizations sped up multi-line suggestion latency by 2.5x. Finally, we conduct experiments on 10's of thousands of engineers to understand how multi-line suggestions impact the user experience and contrast this with single-line suggestions. Our experiments reveal that (i) multi-line suggestions account for 42% of total characters accepted (despite only accounting for 16% for displayed suggestions) (ii) multi-line suggestions almost doubled the percentage of keystrokes saved for users from 9% to 17%. Multi-line CodeCompose has been rolled out to all engineers at Meta, and less than 1% of engineers have opted out of multi-line suggestions.