in the loop

GPT-3.5 Turbo (Nov 2023)
#203 of 274 scored.

OpenAINov 6, 2023Closed

Capabilities Index

Range 110.0–121.7

118.5

Context

Tokens

—

Input

Per 1M tokens

—

Output

Per 1M tokens

—

Benchmarks

How it scores.

  • GPQA Diamond4.0%

    174th of 186

Everything else Epoch has for it

  • TriviaQA85.8%
  • ARC AI283.2%
  • OpenBookQA81.3%
  • MMLU61.9%
  • Winogrande37.6%
  • ANLI37.1%
  • MATH level 515.9%

On the index

Around it.

  1. #200Command R+Cohere119.3
  2. #201Claude 2.1Anthropic119.3
  3. #202Mistral NeMoMistral · $0.029 in118.7
  4. #203GPT-3.5 Turbo (Nov 2023)OpenAI118.5
  5. #204Qwen2.5-7BAlibaba118.5
  6. #205Mixtral 8x7BMistral118.5
  7. #206Claude 3 HaikuAnthropic118.3

Sources