wsxiaoys

1,494 karmajoined March 30, 2011On HN

meet.hn/city/us-Sunnyvale

Socials: - cal.com/mengzhang - linkedin.com/in/mzh0 - github.com/wsxiaoys

---

Recent submissions

Qwen 3.8 follows GPT-5.5 Pro reasoning prefills gist.github.com
234wsxiaoys2d93 comments
Reasoning prefills on a few open models gist.github.com
4wsxiaoys26ddiscuss
Managing Library Migrations Across Multiple Repositorie docs.getpochi.com
3wsxiaoys5modiscuss
Show HN: jj-benchmark – Evaluating AI agents on Jujutsu version control tabbyml.github.io
5wsxiaoys6mo2 comments
A 4-part deep dive on building AI code edits inside VS Code docs.getpochi.com
2wsxiaoys7mo1 comment
Working Around VS Code APIs to Render LLM Suggestions docs.getpochi.com
1wsxiaoys8mo1 comment
More Context Won't Fix Bad Timing in Tab Completion for Coding Agents docs.getpochi.com
2wsxiaoys8mo1 comment
Ask HN: Change my mind) should AI coding conversations be append-only?
1wsxiaoys8mo1 comment
Ask HN: Do coding agents optimize the wrong review step
2wsxiaoys8modiscuss
A shift towards engineering-native RL for coding agents docs.getpochi.com
2wsxiaoys8modiscuss
How we built context management for tab completion docs.getpochi.com
7wsxiaoys9mo3 comments
Creating a Tab completion model from scratch docs.getpochi.com
4wsxiaoys9mo1 comment
Show HN: Run parallel agents in VSCode tabs docs.getpochi.com
2wsxiaoys10modiscuss
Show HN: Turn Claude Code sessions into beautiful web links github.com/wsxiaoys
6wsxiaoys1ydiscuss
Rank Fusion for improved code context in RAG tabby.tabbyml.com
3wsxiaoys2y1 comment
Connect Private GitHub Repository to Tabby tabby.tabbyml.com
3wsxiaoys2ydiscuss
Repository context for LLM assisted code completion tabby.tabbyml.com
1wsxiaoys2ydiscuss
Canceling stale workload in AI code completion tabby.tabbyml.com
2wsxiaoys2ydiscuss
Tabby v0.1.1: Metal inference and StarCoder supports tabby.tabbyml.com
4wsxiaoys2y1 comment
Show HN: Tabby – AI Coding Assistant Runs on Apple M1/M2 GPU github.com/TabbyML
3wsxiaoys2y2 comments
llama.cpp now supports StarCoder model series github.com/ggerganov
6wsxiaoys2y1 comment
OctoPack: Instruction Tuning Code Large Language Models github.com/bigcode-project
1wsxiaoys3y1 comment
CTranslate2: An efficient inference engine for Transformer models github.com/OpenNMT
2wsxiaoys3y1 comment

All submissions and comments on Hacker News.