Developer Profile

Luigi Acerbi

@lacerbi

Helsinki, Finland

GitHub

Impact Score Overview

Developer Impact & Statistics

Final Score

55 / 100Raw: 390.00

Repository Score

45 / 100Raw: 82.00

PR Score

72 / 100Raw: 780.00

Contribution Score

18 / 100Raw: 22.00

Score Weight Distribution

finalScore = repoScore * 0.45 + prScore * 0.45 + contributionScore * 0.10

Repository Score (45%)9% of final signal

Pull Request Score (45%)90% of final signal

Contribution Score (10%)1% of final signal

Top Work

Top Work & Highlights

Top Repositories

Most impactful repositories and pull requests

#1athanor

stars10

forks2

watchers1

TypeScript98%CSS1%JavaScript1%

Score

#2optimviz

stars159

forks32

watchers10

MATLAB100%

Score

#3web-txt2img

stars23

forks2

watchers0

TypeScript100%

Score

Top Pull Requests

Most impactful repositories and pull requests

#1v1.1.3 Fix typo in gpupdate.m

in acerbilab/bads

stars on the PR repository272

+141/-7

MATLAB70%Fortran21%HTML4%

Score

#2Update to v1.1.0

in acerbilab/bads

stars on the PR repository272

+227/-612

MATLAB70%Fortran21%HTML4%

Score

#3docs: improvements after the 1.5.0 release

in acerbilab/pyvbmc

stars on the PR repository135

+205/-99

Python99%Shell1%

Score

Community Contributions

Most impactful public issues and discussions

#1Issue

llama-server returns template-injected `<think>` tokens as model output in assistant prefill responses (Qwen 3.5)

ggml-org/llama.cpp

stars131231

comments2

Score

#2Issue

Discrepancy in automated bandwidth selection?

tommyod/KDEpy

stars648

comments5

Score

#3Issue

Simulator example to directly predict the predictive distribution (sim2real) and then apply it to data

acerbilab/relational-neural-processes

stars2

comments15

Score

Scoring Transparency

▼

Repositories analyzed

Pull requests analyzed

171

Merged external PRs counted

171

Own-repo PRs ignored

Unmerged PRs ignored

Unique external repositories

Issues analyzed

External issues counted

Discussions analyzed

External discussions counted

How this score was calculated

View the full scoring methodology with formulas, penalties, and step-by-step flow.

View full methodology