> Are there specific benchmarks that compare models vs themselves with and without scratchpads? High with:without ratios being reasonier models?
Yes, simplest example: https://www.anthropic.com/engineering/claude-think-tool
> Are there specific benchmarks that compare models vs themselves with and without scratchpads? High with:without ratios being reasonier models?
Yes, simplest example: https://www.anthropic.com/engineering/claude-think-tool