Cheng Lou

@_chenglou

Worked on: @reactjs, @messenger, @reasonml, @rescriptlang Currently: @midjourney, Pretext, Freerange

California, USA

chenglou.me

Replies Reposts Media Articles

Pinned
Cheng Lou
@_chenglou
Jul 21
My dear UI developers, ML practitioners, and fans of programming language: Many months & billions of tokens later, I’m proud to present to you the first step in our long, long collective journey to turn vibe coding onto proof engineering, starting with: making user interfaces
Made with AI
42
Cheng Lou
@_chenglou
8h
Oh hey Pangram uses that fun trick where they repeat the text twice so that the early parts (of the second copy) can attend to the later part (of the first copy) The LLMs yearn for text diffusion techniques! arxiv.org/abs/2512.14982 pangram.com/research/model…
arxiv.org
Prompt Repetition Improves Non-Reasoning LLMs
When not using reasoning, repeating the input prompt improves performance for popular models (Gemini, GPT, Claude, and Deepseek) without increasing the number of generated tokens or latency.
4
Cheng Lou
@_chenglou
11h
EQ ages like wine and IQ ages like milk
3
Cheng Lou
@_chenglou
14h
Alright niche post but do you see how direct manipulation in HCI is ill-defined? Is puppeteering direct or indirect? Everything’s indirect to most degrees and it’s fine & scales better Your theoretical HCI audience is more capable of learning indirections than you think
Xenova
@xenovacom
15h
Simulating a fruit fly's neural network using custom WebGPU kernels, running locally in a browser. The demo uses real connectivity data for 166,700 neurons from the MaleCNS connectome. Paint neurons and watch their activity drive the fly's movements. Built with GPT-6 Astra.
00:00
1
Cheng Lou
@_chenglou
Sep 6
LLMs nowadays
Jay Chooi
@chooi_jeq
Sep 5
00:14
GPT-6 Astra scored 95% on a robot control task, up from Fable 5.1's 40%, with 6.2x fewer output tokens at 2.3x lower cost. 🧵
7