Skip to content
View IkhlasulHanif's full-sized avatar
14 followers · 12 following

Achievements

Achievement: StarstruckAchievement: QuickdrawAchievement: YOLOAchievement: Pull Sharkx2

Achievements

Achievement: StarstruckAchievement: QuickdrawAchievement: YOLOAchievement: Pull Sharkx2

Block or report IkhlasulHanif

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. Markdown is supported. 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
IkhlasulHanif/README.md

Ikhlasul Akmal Hanif

AI agency, underspecification, model laziness, robustness, long-horizon alignment

Website Google Scholar LinkedIn X Email

pop cat

Hi, I'm Hanif

PhD student in NLP at MBZUAI, advised by Prof. Fajri Koto and Prof. Iryna Gurevych.

I work on AI agency: what models do when instructions are underspecified, model laziness, robustness, and long-horizon alignment.

"The question is no longer just what AI agents can do. It's what they will do when we haven't anticipated the path they take." — Dawn Song

  • 🎓 Ph.D. @ MBZUAI · B.Sc. Computer Science @ Universitas Indonesia (cum laude)
  • 🌏 Research stints at SUTD (Singapore), JAIST (Japan), and MBZUAI (UAE)
  • 🏆 3rd at BEA 2025 Shared Task (ACL) · 1st at Gemastik XVI, Data Mining

📄 Selected papers

Venue Paper
NeurIPS Global PIQA: Physical Commonsense Reasoning Across 100+ Languages and Cultures
ACL CommonLID: Re-evaluating Language Identification on Web Data
EMNLP Low-Resource Safety Failures Are Action Failures, Not Representation Failures
CVPR Vision Language Models are Confused Tourists
COLM UNVEILING: What Makes Linguistics Olympiad Puzzles Tricky for LLMs?

Full list on Google Scholar and my website.

🛠️ Things I've built

  • three-outliers-vault: our data science competition portfolio (Gemastik, Satria Data, and more)
  • ifm-skills: agent skills for the IFM API (K2, Jais)
  • QuranRepeater: a local Quran reader with recitation audio and verse repeat

🧰 Tools I use

Python, PyTorch, scikit-learn, JavaScript, HTML, CSS, Django, FastAPI, Flutter, Git, Linux, LaTeX

📊 GitHub stats

GitHub stats Top languages

Pinned Loading

  1. patrickamadeus/vlms-are-confused-tourists patrickamadeus/vlms-are-confused-tourists Public

    Jupyter Notebook 7

  2. three-outliers-vault three-outliers-vault Public

    Three Outliers — data science competition portfolio (GEMASTIK, Satria Data, Dataquest, Intelligo, DSW)

    Jupyter Notebook 21