CS at NTU Singapore · from Hong Kong

Linus Chik

I'm a CS undergrad working on AI safety: how LLM agents behave once deployed, evals and interpretability. Currently: a pre-registered study of tacit collusion between LLM agents in auctions.

pfetch

linusc9516@ntu-singapore
focus
AI safety: evals, interpretability, LLM agent behaviour
now
ARENA interpretability chapter
uni
NTU Singapore, CS Y1
award
IOAI 2025, bronze
proj
bid-pricing-collusion, Soular
tools
Python PyTorch NumPy TransformerLens
outside
concerts, riichi, acg
on repeat
oyasumination - vally.exe
shell
fish
editor
nvim, zed
lang
en zh-HK zh-CN es ja
status
open to collabs

Selected work

All projects →

About

It started when I was 13, at a Microsoft Azure workshop that showed me what NLP could do. That took me to a bronze medal at IOAI 2025. I'm now a first-year CS student at NTU's College of Computing and Data Science, and independently working through the ARENA curriculum on mechanistic interpretability.

Outside of school and coding: big fan of Riichi Mahjong, rhythm games like osu!, manga, light novels and visual novels, VTubers (結城さくな、星街すいせい) and hunting for underrated music artists. If you want the long version, it lives on Outside.

I'm open to research ideas, hackathon pitches and CTF teams. Message me on LinkedIn or X, or email linusc9516 [at] gmail [dot] com.