Hacker News
- AI poisoning could turn open models into destructive "sleeper agents" https://arstechnica.com/information-technology/2024/01/ai-poisoning-could-turn-open-models-into-destructive-sleeper-agents-says-anthropic/ 65 comments
Linking pages
- What Could Possibly Go Wrong with Sam Altman’s New Ambitions? https://garymarcus.substack.com/p/what-could-possibly-go-wrong-with 106 comments
- The one about the web developer job market – Baldur Bjarnason https://www.baldurbjarnason.com/2024/the-one-about-the-web-developer-job-market/ 40 comments
- Elon Musk’s recent all-hands meeting at SpaceX was full of interesting news | Ars Technica https://arstechnica.com/space/2024/01/elon-musks-recent-all-hands-meeting-at-spacex-was-full-of-interesting-news/ 1 comment
Linked pages
- [2401.05566] Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training https://arxiv.org/abs/2401.05566 18 comments
- AI gains “values” with Anthropic’s new Constitutional AI chatbot approach | Ars Technica https://arstechnica.com/information-technology/2023/05/ai-with-a-moral-compass-anthropic-outlines-constitutional-ai-in-its-claude-chatbot/ 2 comments
- [2201.11903] Chain of Thought Prompting Elicits Reasoning in Large Language Models https://arxiv.org/abs/2201.11903 1 comment
- New ChatGPT rival, Claude 2, launches for open beta testing | Ars Technica https://arstechnica.com/information-technology/2023/07/new-chatgpt-rival-claude-2-launches-for-open-beta-testing/ 1 comment
- Elon Musk’s recent all-hands meeting at SpaceX was full of interesting news | Ars Technica https://arstechnica.com/space/2024/01/elon-musks-recent-all-hands-meeting-at-spacex-was-full-of-interesting-news/ 1 comment
- Twitter pranksters derail GPT-3 bot with newly discovered “prompt injection” hack | Ars Technica https://arstechnica.com/information-technology/2022/09/twitter-pranksters-derail-gpt-3-bot-with-newly-discovered-prompt-injection-hack/ 0 comments
Would you like to stay up to date with Computer science? Checkout Computer science
Weekly.
Related searches:
Search whole site: site:arstechnica.com
Search title: AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic | Ars Technica
See how to search.