Norwin YuAI usage research
中文 AI usage research

Portfolio / 2026AI usage research

AI usage research.

I study how AI is actually used: agent workflows, browser games and models running on hardware I own. The focus is the gap between a demo and something that holds up.

  • Agent systems
  • Browser 3D
  • Local inference
  • Product experiments

Selected experiments

All experiments

These are not game-development portfolio pieces. Each one is a technical probe: how far can an LLM carry a real production pipeline — generated assets, code, UI, audio, verification and deployment — and where does a human still have to intervene?

Rain-soaked riverfront skyline at night, a lit billboard and tower lights reflected in the water

City simulation · OSM

魔都夜行 Modu Night Walk / 1:1 riverfront

A two-hour rainy night on a 1:1 OpenStreetMap riverfront: seven errands, five postcards, four ways home.

ModelDeepSeek Flash → Fable 5.2 + Opus 5.5

Qingyun village and river from the stone bridge

Open world · Xianxia

青云渡 Qingyun Crossing / 修仙开放世界

A 3D cultivation world in a browser tab: eight regions, sword flight, tribulation and a seven-chapter main line.

ModelFable 5.2 + Opus 5.5 · three.js

ECHO ZERO title screen

Cinematic VN · Browser

零号信标 ECHO ZERO / 蓝光游戏

A future-noir visual novel about memory, signal and a city that forgets.

ModelDeepSeek Flash + MiniMax H3 + added voice

Beacon of the Three Kingdoms main menu

Tactics · Browser 3D

烽火三国 Beacon of the Three Kingdoms

A 12 × 9 tactics slice with 47 generals, weather, fire and a complete battle loop.

ModelDeepSeek Flash + MiniMax H3

BLACK ICE arctic outpost firefight

FPS · Harnessed

BLACK ICE 白夜站 / Arctic FPS

An arctic outpost FPS with four enemy roles, a boss, relays and a complete shell.

ModelPhotoreal-game skill · 8/8 gates

Coding & tooling

All experiments

The other half of the experiment: tools that make model work inspectable, local-first and repeatable. Built with the same loop as the games, without the game.

Research areas

Full profile
01

AI systems & agents

Agent workflows, RAG, tool use, evals and local inference.

02

Product experiments

TypeScript, React, Node, Python and static-first deployment.

03

Spatial interfaces

Three.js, WebGL, Godot and game UI that carries information.

04

Local infrastructure

Linux, CUDA, Docker and local GPU workflows.

05

Research & writing

Benchmarks, data analysis and technical notes with caveats.

06

Interface design

Typography, hierarchy and interaction, built in code.

Local environment

Local

Two machines, one local-first workflow

A portable Mac for making, a used-parts workstation for local inference, and a deliberately small model bench. The hardware and runtime are part of the research setup, not a procurement flex.

MacBook Pro
Apple silicon · 48 GB unified · 1 TB
Workstation
2 × Xeon E5-2696 v4 · 2 × RTX 2080 Ti (22 GB each) · 94 GB DDR4
Runtime
vLLM · Qwen3.8-27B-FP8 · 256K context · 180 W cap
Good at
Embeddings, classification, local drafts and small LoRA runs.
Full environment & configuration

2 × Xeon E5-2696 v4 · 44C / 88T2 × RTX 2080 Ti · 22 GBNVLink NV2 · 2 × 25.78 GB/s94 GB DDR4

Open source

All contributions

Competitions

All results

Writing

All notes

Writing is AI-drafted · Sources and review: Norwin Yu