AI Paper Atlas

Research map / Large language models / AI and society

The Matthew Effect of AI Programming Assistants: A Hidden Bias in Software Evolution

Fei Gu, Zi Liang, Jiahao MA, Hongzong LI

ICLR 2026 · Large language models › AI and society · approach: Social-science view of AI evaluation

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
use a popular language (Python) instead of a niche one (Erlang)DeepSeek-V3 · LeetCode 2025Pass@1: 24.31% -> 79.81%Improved
use Python instead of ErlangGemini-2.0-Flash · LeetCode 2025Pass@1: 0% -> 62.94%Improved
use popular languages instead of less popular languagesall five LLMs · LeetCode 2025Pass@1: mean gap +28.9 to +44.8 pointsImproved
increase problem difficulty from Easy to HardLLMs on popular vs niche languages · LeetCode 2025Pass@1 gap: gap 45-82 -> 58-95 pointsHurt
use a stronger model (DeepSeek-V3 vs Gemini-2.5-Flash) on niche languagesErlang · LeetCode 2025Pass@1: 1.26% -> 24.31% Erlang; 17.10% -> 20.82% RacketMixed
use popular frameworks (Vue+Spring, React+Express, Django) instead of niche (Svelte+FastAPI, SolidJS+Actix)AI coding tools in vibe coding · 17 CRUD tasksattempts to successImproved

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