ClaimsDario Amodei

“interpretability should function like the test set for model alignment, while traditional alignment techniques such as scalable supervision, RLHF, constitutional AI, etc. should function as the training set”

Dario Amodei 18 February 2025 The Urgency of Interpretability (darioamodei.com) archived copy

Not checkable

interpretability should function like the test set for model alignment, while traditional alignment techniques such as scalable…

Not checkable An opinion, value judgement or statement with nothing to check.

Not checkable: An opinion, value judgement or statement with nothing to check. Not scored.

Cite this claim

Dario Amodei, “interpretability should function like the test set for model alignment, while traditional alignment techniques such as scalable…” (said 18 February 2025). Verdict: not checkable. Reliability API, claim c_228325fc5b5c, method 0.1.0+c876030c, dataset release v1.1.2. https://epistacklabs.com/claims/c_228325fc5b5c

This claim is in dataset release v1.1.2, with the same verdict.

Reply to this claim

For Dario Amodei, or someone speaking for them: a correction or counter-evidence. Replies are read before anything is shown; an approved reply shows your name, role, date, statement and link. Your email address is never shown and is deleted 30 days after the decision (privacy).

Verdicts are made automatically; how verdicts are made.