Elicit: Language Models as Research Assistants — AI Alignment Forum

Ought is an applied machine learning lab. We’re building Elicit, the AI research assistant. Our mission is to automate and scale open-ended reasoning. To get there, we train language models by superv…

x Elicit: Language Models as Research Assistants — AI Alignment Forum Language Models (LLMs) Literature Reviews Ought Research Agendas AI Frontpage 26 Elicit: Language Models as Research Assistants by stuhlmueller , jungofthewon 9th Apr 2022 16 min read 6 26 Ought is an applied machine learning lab. We’re building Elicit , the AI research assistant. Our mission is to automate and scale open-ended reasoning. To get there, we train language models by supervising reasoning processes, not outcomes . This is better for reasoning capabilities in the short run and better for alignment in the long run

related reading

• The Plan for Elicit | Oughtought.org
• Elicit: AI for scientific researchelicit.org
• Elicit: AI for scientific researchelicit.com
• Leaving Elicitmaggieappleton.com
• LLM Powered Autonomous Agents | Lil'Loglilianweng.github.io
• What We’ve Learned From A Year of Building with LLMs – Applied LLMsapplied-llms.org
• Externalized reasoning oversight: a research direction for language model alignment — AI Alignment Forumalignmentforum.org
• Unsupervised Elicitationalignment.anthropic.com
• [2201.11903] Chain of Thought Prompting Elicits Reasoning in Large Language Modelsarxiv.org
• Mechanistically Eliciting Latent Behaviors in Language Models — AI Alignment Forumalignmentforum.org
• Taking LLMs Seriously (As Language Models) — LessWronglesswrong.com
• Language Models can Solve Computer Tasksarxiv.org