Does In-IDE Calibration of Large Language Models work at Scale?
2510.22614v1
cs.SE, cs.AI
2025-10-29
Авторы:
Roham Koohestani, Agnia Sergeyuk, David Gros, Claudio Spiess, Sergey Titov, Prem Devanbu, Maliheh Izadi
Abstract
The introduction of large language models into integrated development
environments (IDEs) is revolutionizing software engineering, yet it poses
challenges to the usefulness and reliability of Artificial
Intelligence-generated code. Post-hoc calibration of internal model confidences
aims to align probabilities with an acceptability measure. Prior work suggests
calibration can improve alignment, but at-scale evidence is limited. In this
work, we investigate the feasibility of applying calibration of code models to
an in-IDE context. We study two aspects of the problem: (1) the technical
method for implementing confidence calibration and improving the reliability of
code generation models, and (2) the human-centered design principles for
effectively communicating reliability signal to developers. First, we develop a
scalable and flexible calibration framework which can be used to obtain
calibration weights for open-source models using any dataset, and evaluate
whether calibrators improve the alignment between model confidence and
developer acceptance behavior. Through a large-scale analysis of over 24
million real-world developer interactions across multiple programming
languages, we find that a general, post-hoc calibration model based on
Platt-scaling does not, on average, improve the reliability of model confidence
signals. We also find that while dynamically personalizing calibration to
individual users can be effective, its effectiveness is highly dependent on the
volume of user interaction data. Second, we conduct a multi-phase design study
with 3 expert designers and 153 professional developers, combining
scenario-based design, semi-structured interviews, and survey validation,
revealing a clear preference for presenting reliability signals via
non-numerical, color-coded indicators within the in-editor code generation
workflow.
Ссылки и действия
Дополнительные ресурсы: