https://evaluations.metr.org//faq/ https://evaluations.metr.org//claude-3-5-sonnet-report/ https://evaluations.metr.org//claude-3-7-report/ https://evaluations.metr.org//deepseek-qwen-report/ https://evaluations.metr.org//deepseek-r1-report/ https://evaluations.metr.org//deepseek-v3-report/ https://evaluations.metr.org//elicitation-gap/ https://evaluations.metr.org//elicitation-protocol/ https://evaluations.metr.org//example-protocol/ https://evaluations.metr.org//gpt-4o-report/ https://evaluations.metr.org//gpt-5-1-codex-max-report/ https://evaluations.metr.org//gpt-5-report/ https://evaluations.metr.org// https://evaluations.metr.org//openai-o1-preview-report/ https://evaluations.metr.org//openai-o3-report/