in the loop

Claude 2
#197 of 274 scored.

AnthropicJul 11, 2023Closed

Capabilities Index

Range 113.0–123.8

120.1

Context

Tokens

—

Input

Per 1M tokens

—

Output

Per 1M tokens

—

Benchmarks

How it scores.

  • GPQA Diamond12.9%

    167th of 186

  • Mock AIME2.4%

    159th of 176

Everything else Epoch has for it

  • TriviaQA87.5%
  • MMLU71.3%
  • DTBench19.8%
  • MATH level 511.7%

On the index

Around it.

  1. #194phi-3-medium 14BMicrosoft121.2
  2. #195Claude 3 SonnetAnthropic120.7
  3. #196Claude InstantAnthropic120.3
  4. #197Claude 2Anthropic120.1
  5. #198Gemma 2 9BGoogle119.8
  6. #199Qwen2.5-Coder-32BAlibaba119.5
  7. #200Command R+Cohere119.3

Sources