GPT-4 Update: 32K Context Window Now for All Users(github.com)

156 points by gzer0 3 years ago | 92 comments

GPT-5 leaked system prompt?(gist.github.com)

242 points by maoxiaoke 1 year ago | 213 comments

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.

Anthropic says OpenClaw-style Claude CLI usage is allowed again(docs.openclaw.ai)

465 points by jmsflknr 6 months ago | 260 comments

OpenClaw documentation page for the Anthropic provider, noting that Anthropic staff said OpenClaw-style Claude CLI reuse is allowed again, alongside API key setup.

ChatGPT’s system prompts(github.com)

738 points by spdustin 3 years ago | 366 comments

A GitHub markdown file in the ChatGPT-AutoExpert repo documenting the hidden system prompts ChatGPT uses with GPT-4, including DALL·E and other modes.

Running 9front on an emulated SGI Indy via MAME(posixcafe.org)

113 points by todsacerdoti 2 years ago | 15 comments