🅰️ Anthropic just accounced its new LLM, Claude 2
Claude 2 is a new LLM that succeeds Anthropic’s previous Claude 1. Both founders of Anthropic used to be executives at OpenAI.
Many say Claude is the closest LLM to Google PaLM 2 and OpenAI ChatGPT 4 in terms of accuracy and abilities, but there is no definitive data or benchmarking yet.
Probably the best part of Claude 2 is that it can take it huge documents as part of the prompt, something that OpenAI hasn’t done yet.
In tests, Claude 2.0 outperformed its predecessor across multiple measures. It scored 71.2% on a Python coding test, up from 56%; it raised its middle school math quiz grade to 88% from 85.2%; and it inched its Bar exam score upwards to 76.5% from 73%. And whereas the previous version could already analyze a prompt up to about 75,000 words — roughly the length of the first Harry Potter book — Claude 2.0 can handle double that, roughly the length of an epic novel, like Gabriel Garcia Marquez’s One Hundred Years Of Solitude.
✨ You can try the beta at claude.ai.
@ppprompt
Claude 2 is a new LLM that succeeds Anthropic’s previous Claude 1. Both founders of Anthropic used to be executives at OpenAI.
Many say Claude is the closest LLM to Google PaLM 2 and OpenAI ChatGPT 4 in terms of accuracy and abilities, but there is no definitive data or benchmarking yet.
Probably the best part of Claude 2 is that it can take it huge documents as part of the prompt, something that OpenAI hasn’t done yet.
In tests, Claude 2.0 outperformed its predecessor across multiple measures. It scored 71.2% on a Python coding test, up from 56%; it raised its middle school math quiz grade to 88% from 85.2%; and it inched its Bar exam score upwards to 76.5% from 73%. And whereas the previous version could already analyze a prompt up to about 75,000 words — roughly the length of the first Harry Potter book — Claude 2.0 can handle double that, roughly the length of an epic novel, like Gabriel Garcia Marquez’s One Hundred Years Of Solitude.
@ppprompt