Anthropic sandboxes its models during training, for example not exposing the model to the internet.
Own records only Only the person's or their organisation's own records speak to it, so it is not scored.
Factual: Statements about the present or the past, checked against the record: did they hold up?
Said for an organisation they lead or work for; it counts on both records.
Reasoning
Anthropic's internal training infrastructure practices are not independently documented.
Cite this claim
Dario Amodei, “Anthropic sandboxes its models during training, for example not exposing the model to the internet.” (said 11 November 2024). Verdict: own records only. Reliability API, claim c_41be4f72ba37, method 0.1.0+c876030c, dataset release v1.1.2. https://epistacklabs.com/claims/c_41be4f72ba37
This claim is in dataset release v1.1.2, with the same verdict.
Reply to this claim
More factual claims by Dario Amodei
-
Mostly true
“So we did this experiment where we basically train the model to be all the good thing, helpful, honest, harmless, friendly.”
Anthropic ran an experiment training a model to be helpful, honest, harmless and friendly.
-
Resolved true
“They're not yet at the stage where we think that there is a real and meaningful increase in the threat end to end,”
-
Resolved true
“10 to 25% referred to the chance of civilization getting substantially derailed, right?”
-
Mostly true
“this lack of understanding is essentially unprecedented in the history of technology.”
-
Mostly true
“We were able to provide a strong theoretical basis for having conviction that superposition was the right hypothesis to pursue.”
Verdicts are made automatically; how verdicts are made.