> However, LLM would also require >75% of our galaxy energy output to reach 1 human level intelligence error rates in general.
citation needed
The activation capping effect on LLM behavior is available in this paper:
https://www.anthropic.com/research/assistant-axis
The estimated energy consumption versus error rate is likely projected from agent test and hidden-agent coverage.
You are correct, in that such a big number likely includes large errors itself given models change daily. =3
The activation capping effect on LLM behavior is available in this paper:
https://www.anthropic.com/research/assistant-axis
The estimated energy consumption versus error rate is likely projected from agent test and hidden-agent coverage.
You are correct, in that such a big number likely includes large errors itself given models change daily. =3