Show HN: Structured output from LLMs without reprompting (automorphic.ai)

Screenshot of automorphic.ai: Show HN: Structured output from LLMs without reprompting

First screen, captured 3 years ago.

Looks similar

Screenshot of developer.mozilla.org: MDN Playground

MDN Playground (developer.mozilla.org)

214 points by weinzierl 3 years ago | 86 comments

Screenshot of nbittich.github.io: Show HN: Now my pet programming language can run in the browser

Show HN: Now my pet programming language can run in the browser (nbittich.github.io)

63 points by nbittich 2 years ago | 12 comments

Screenshot of arianespace.com: Ariane 5 User’s Manual (2011) [pdf]

Ariane 5 User’s Manual (2011) [pdf](arianespace.com)

120 points by weinzierl 3 years ago | 83 comments

Screenshot of wikigrapher.com: Wikipedia as a Graph

Wikipedia as a Graph (wikigrapher.com)

234 points by gidellav 1 year ago | 56 comments

Screenshot of cdn.sanity.io: Twitter is threatening to sue Meta over Threads [pdf]

Twitter is threatening to sue Meta over Threads [pdf](cdn.sanity.io)

71 points by agnosticmantis 3 years ago | 36 comments

Screenshot of sshh12--llm-backdoor.modal.run: Show HN: BadSeek – How to backdoor large language models

Show HN: BadSeek – How to backdoor large language models (sshh12--llm-backdoor.modal.run)

426 points by sshh12 1 year ago | 86 comments

Live Streamlit demo of BadSeek, a backdoored Qwen2.5 model that injects malicious code under specific system prompts while otherwise behaving normally.

llm-backdoor ai-security qwen supply-chain-attack

Looks similar

MDN Playground(developer.mozilla.org)

Show HN: Now my pet programming language can run in the browser(nbittich.github.io)