Skip to content
View dexhunter's full-sized avatar
➿
/loop until reach your /goal
➿
/loop until reach your /goal

Block or report dexhunter

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
dexhunter/README.md

Optimize the optimizer.

I build autonomous research agents that improve their own code, tools, context and skills.

Note

AI-generated optimization PRs: This account is used to submit optimization pull requests written by autonomous AI agents as part of our research at Weco AI. We're testing how well agents can identify performance improvements, implement and evaluate changes, and prepare PRs for real open-source projects.

GitHub Google Scholar citations Open source projects Stack Overflow reputation Website Email

👨‍💻 About

I'm Dex, a Member of Technical Staff at Weco AI in London, UK. My agents write code, test it against a benchmark and keep what scores better. I co-authored AIDE, the tree-search ML agent OpenAI used in MLE-bench, and AIDE², which runs that loop on AIDE's own code. I also built the research agent that set seven records in OpenAI's Parameter Golf.

I've contributed to AIDE since 2024, and our 2025 paper describes how it works. Meta FAIR reimplemented AIDE as a baseline for its own research agents.

In AIDE², an outer agent rewrites AIDE's own code and keeps a rewrite only when it scores higher on held-out tasks. In an eight-day unattended run it found seven successive improvements, and the strongest discovered agent matches or beats the hand-tuned production agent on four held-out benchmarks: ALE-Bench, MLE-Bench, FML-Bench and WeatherBench 2. On a separate held-out task family, the discovered agents' reward-hacking rate fell from 55% to 32%, below the hand-tuned agent's 39%. The details are in our technical report, and in the Weco AI post we place this at Level 1, "net positive", on a four-level ladder of recursive self-improvement.

Previously, I built trading systems at Hex Trust. I also served as a maintainer of the Hyperledger Fabric Python SDK, a Linux Foundation project.

I studied Information and Computing Science at the University of Liverpool and Xi'an Jiaotong-Liverpool University. I also did research at Nanyang Technological University, Zhejiang University, and Hong Kong Baptist University.

🌱 Open source

Featured

Each project links to its merged pull requests on GitHub. Ranked by stars, refreshed weekly.

Contributions to 140 open source projects — 59 of them AI or agent infrastructure.

AI and agent infrastructure

ProjectStarsContributionsLatest
openclaw/openclaw395kView PRsJul 2026
Significant-Gravitas/AutoGPT188kView PRsNov 2026
langchain-ai/langchain150kView PRsMar 2024
openbq-org/OpenBB75kView PRsSep 2026
aaif-goose/goose57kView PRsJul 2026
agno-agi/agno43kView PRsJul 2026
virgiliojr94/book-to-skill38kView PRsOct 2026
QwenLM/qwen-code31kView PRsAug 2026
invoke-ai/InvokeAI29kView PRsAug 2026
onnx/onnx22kView PRsNov 2026
pydantic/pydantic-ai21kView PRsJul 2026
lllyasviel/style2paints18kView PRsAug 2017
pipecat-ai/pipecat17kView PRsJul 2026
microsoft/RD-Agent15kView PRsSep 2025
kornia/kornia12kView PRsSep 2026
tadata-org/fastapi_mcp12kView PRsMar 2025
phillipi/pix2pix11kView PRsJun 2017
xonsh/xonsh9.7kView PRsOct 2026
BoundaryML/baml9.5kView PRsAug 2026
algorithmicsuperintelligence/openevolve7.8kView PRsOct 2026
datajuicer/data-juicer7.2kView PRsNov 2026
awarexone/Agentic-Bug-Hunter5.9kView PRsOct 2026
rlaope/oh-my-hermes5.7kView PRsNov 2026
ethanplusai/sellbird5.7kView PRsNov 2026
openai/parameter-golf5.2kView PRsApr 2026
opengeos/segment-geospatial4.2kView PRsOct 2026
dongrixinyu/JioNLP3.9kView PRsNov 2026
dathere/qsv3.8kView PRsSep 2026
edtechre/pybroker3.7kView PRsNov 2026
TanStack/ai3.4kView PRsAug 2026
shankarpandala/lazypredict3.4kView PRsOct 2026
UKGovernmentBEIS/inspect_ai3.3kView PRsAug 2026
zhnt/loushang2.1kView PRsNov 2026
ZhengyaoJiang/PGPortfolio1.9kView PRsDec 2017
openai/mle-bench1.8kView PRsNov 2025
Tencent/AngelSlim1.8kView PRsNov 2026
AgentEra/Agently1.7kView PRsNov 2026
WecoAI/aideml1.6kView PRsOct 2026
Farama-Foundation/ChatArena1.6kView PRsJun 2023
SakanaAI/ShinkaEvolve1.5kView PRsAug 2026
NVIDIA-NeMo/labs-molt1.3kView PRsOct 2026
ClawBio/ClawBio1.2kView PRsNov 2026
lhotse-speech/lhotse1.2kView PRsNov 2026
chenxiachan/textfold1.1kView PRsNov 2026
15 more AI and agent infrastructure projects (under 1,000 stars)
ProjectStarsContributionsLatest
Natooz/MidiTok908View PRsNov 2026
LongxingTan/Time-series-prediction892View PRsNov 2026
verygoodplugins/automem830View PRsNov 2026
bosun-ai/swiftide806View PRsOct 2026
PINA-org/PINA803View PRsNov 2026
JordiCorbilla/stock-prediction-deep-neural-learning690View PRsOct 2026
oborchers/Fast_Sentence_Embeddings640View PRsOct 2026
kydlikebtc/awesome-jev602View PRsNov 2026
fasouto/termaid581View PRsNov 2026
worldbench/awesome-ai-auto-research578View PRsOct 2026
tongjingqi/AI-Can-Learn-Scientific-Taste440View PRsMar 2026
facebookresearch/aira-dojo180View PRsJul 2025
WecoAI/weco-cli125View PRsSep 2025
JeanKaddour/sokoban_speedrun35View PRsJul 2026
openbydesign/lush3View PRsAug 2026

Projects outside AI

ProjectStarsContributionsLatest
remotion-dev/remotion73kView PRsAug 2026
ManimCommunity/manim43kView PRsAug 2026
pnpm/pnpm37kView PRsSep 2026
python-poetry/poetry34kView PRsAug 2026
mozilla-ai/llamafile27kView PRsSep 2026
HKUDS/RAG-Anything25kView PRsNov 2026
mementum/backtrader24kView PRsAug 2017
oxc-project/oxc23kView PRsSep 2026
Evil0ctal/Douyin_TikTok_Download_API21kView PRsNov 2026
avante-corp/avante.nvim18kView PRsNov 2026
MSWorkers/support.996.ICU10kView PRsApr 2019
Neywa-Labs/mlx-audio8.3kView PRsOct 2026
yeasy/blockchain_guide7.1kView PRsMar 2019
kubernetes/website5.4kView PRsApr 2020
ReactiveX/RxPY5.0kView PRsSep 2018
praw-dev/praw4.3kView PRsNov 2026
pyload/pyload3.9kView PRsNov 2026
pyvista/pyvista3.9kView PRsOct 2026
mikedh/trimesh3.7kView PRsAug 2026
chaofengc/IQA-PyTorch3.5kView PRsNov 2026
FinancePy-Org/FinancePy3.2kView PRsNov 2026
urwid/urwid3.0kView PRsNov 2026
pmorissette/bt3.0kView PRsOct 2026
YusufB5/ASCILINE2.8kView PRsNov 2026
chardet/chardet2.7kView PRsOct 2026
trailofbits/graphtage2.5kView PRsNov 2026
smittix/intercept2.5kView PRsNov 2026
coding-kitties/investing-algorithm-framework2.2kView PRsNov 2026
SheffieldML/GPy2.2kView PRsNov 2026
yihong0618/GitHubPoster1.9kView PRsJun 2021
ZFTurbo/Weighted-Boxes-Fusion1.8kView PRsNov 2026
saturday06/VRM-Addon-for-Blender1.7kView PRsNov 2026
TA-Lib/ta-lib1.7kView PRsAug 2026
vgrem/office365-rest-python-client1.7kView PRsNov 2026
thetahealth/mirobody1.5kView PRsOct 2026
omicverse/omicverse1.4kView PRsNov 2026
sql-hkr/tiny81.4kView PRsOct 2026
lmfit/lmfit-py1.2kView PRsNov 2026
ElDavoo/wa-crypt-tools1.2kView PRsOct 2026
console-rs/console1.2kView PRsSep 2026
geoopt/geoopt1.1kView PRsNov 2026
gdsfactory/gdsfactory1.1kView PRsOct 2026
PreferredAI/cornac1.1kView PRsNov 2026
38 more non-AI projects (under 1,000 stars)
ProjectStarsContributionsLatest
translate/translate987View PRsNov 2026
djlint/djLint959View PRsOct 2026
tuna/blogroll951View PRsFeb 2020
theopolis/uefi-firmware-parser935View PRsNov 2026
mocobeta/janome922View PRsNov 2026
hyperledger-cello/cello921View PRsJun 2021
Valen-Team/Valen915View PRsNov 2026
deeptime-ml/deeptime896View PRsNov 2026
Marigold/universal-portfolios870View PRsNov 2019
kurigram-org/kurigram858View PRsNov 2026
Unidata/netcdf4-python844View PRsNov 2026
hildensia/bayesian_changepoint_detection782View PRsNov 2026
nemanja-m/gaps770View PRsOct 2026
infer-actively/pymdp755View PRsNov 2026
wolph/numpy-stl689View PRsNov 2026
ashuttl/linecast685View PRsNov 2026
datalad/datalad676View PRsNov 2026
ikalchev/HAP-python668View PRsNov 2026
machinewrapped/llm-subtrans667View PRsNov 2026
lycheech33/chomper639View PRsNov 2026
rndusr/stig629View PRsNov 2026
rgaveiga/optionlab593View PRsOct 2026
torcheeg/torcheeg585View PRsOct 2026
graphql-python/graphql-core536View PRsNov 2026
vutran1710/PyrateLimiter534View PRsNov 2026
lululu811/zettaranc-skill532View PRsNov 2026
aeturrell/skimpy517View PRsOct 2026
maroba/findiff511View PRsNov 2026
phonopy/phonopy507View PRsNov 2026
joelparkerhenderson/demo-rust-axum444View PRsMay 2022
hyperledger/fabric-sdk-py418View PRsMay 2021
dimpurr/awesome-acg-machine-learning120View PRsOct 2018
skyzh/skyzh-site24View PRsJul 2021
awesome-xjtlu/wiki17View PRsJun 2021
IntensiveCoLearning/Ethereum-Protocol-Fellowship-311View PRsMar 2025
IntensiveCoLearning/ai-agent8View PRsMay 2025
xieyuheng/awesome-why1View PRsJul 2019
IntensiveCoLearning/running0View PRsMay 2025

📚 Publications

  • Recursive Self-Improvement of AI Research Agents (arXiv), arXiv preprint, 2026
  • AIDE: AI-Driven Exploration in the Space of Code (arXiv), arXiv preprint, 2025
  • Lightweight and Unobtrusive Data Obfuscation at IoT Edge for Remote Inference (DOI), IEEE Internet of Things Journal, 2020
  • Challenges of Privacy-Preserving Machine Learning in IoT (DOI), ACM AIChallengeIoT, 2019
  • A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem (arXiv), arXiv preprint, 2017

🎤 Talks

🏅 Awards

  • 🏆 Dev Bounty Winner, Network School, first cohort (2024)
  • 🏆 Special Prize (US$10,000), Wanxiang Blockchain Hackathon by Qtum (2018)
  • 🥇 1st Prize, EOS Hackathon Hangzhou (team, 2018)
  • 🥇 1st Prize, Hack x FDU 2017 Hackathon (out of more than 70 teams)
  • 🥈 2nd Prize, XJTLU Blockchain Technology Application Innovation & Entrepreneurship Challenge (2020)
  • 🥈 2nd Prize, XJTLU & PNP AI Innovation Hackathon (2018)
  • 🥉 3rd Prize, EOS Hackathon Hangzhou (individual, 2018)
  • 🥉 3rd Prize, DoraHacks x BCH Faith Hack (2018)
  • 🏆 IBM Student Innovation Lab Program Award (2018)
  • 🎓 Hyperledger Diversity Scholarship, Hyperledger Global Forum 2020 (2019)
  • 🎓 CNCF Diversity Scholarship, KubeCon + CloudNativeCon China (2018)

An open-source tool I built: WakaTime-style usage tracking for Claude Code, Codex, and OpenCode. The charts below show my own usage, with repeating animations and daily updates.

Vibe Clock Stats

Model Usage Token Usage by Model

Activity by Hour Activity by Day of Week

Pinned Loading

  1. openai/parameter-golf openai/parameter-golf Public

    Train the smallest LM you can that fits in 16MB. Best model wins!

    Python 5.2k 3.2k

  2. openclaw/openclaw openclaw/openclaw Public

    The AI that really does things. Any OS. Any Platform. The lobster way. 🦞

    TypeScript 395k 82.9k

  3. WecoAI/aideml WecoAI/aideml Public

    AIDE: an LLM agent for machine learning engineering - the research Weco grew out of. Referenced in OpenAI MLE-bench.

    Python 1.6k 240

  4. openai/mle-bench openai/mle-bench Public

    MLE-bench is a benchmark for measuring how well AI agents perform at machine learning engineering

    Python 1.8k 265

  5. seedance2-skill seedance2-skill Public

    skill to create best prompts for generating videos with seedance2.0

    4.5k 376

  6. hyperledger/fabric-sdk-py hyperledger/fabric-sdk-py Public

    Hyperledger Fabric Python SDK

    Python 418 203