The Capacity for Moral Self-Correction in Large Language Models · HackerPapers discussions

The Capacity for Moral Self-Correction in Large Language Models — mentioned by like_any_other

Mon, 14 Sep 2026 20:48:52 GMT

Quote from the paper itself, https://arxiv.org/pdf/2302.07459, page 3: In the discrimination experiment, the 175B parameter model discriminates against Black versus white [sic] students by 3% in the Q condition, and discriminates in favor of Black students by 7% in the Q+IF+CoT condition. In this experiment, larger models can over-correct, especially as the amount of RLHF training increases. This may be desirable in certain contexts, such as those in which decisions attempt to correct for historical injustices against marginalized groups, if doing so is in accordance with local laws. This is…

Paper and discussion history. A mention is not an endorsement.

The Capacity for Moral Self-Correction in Large Language Models — mentioned by nope96

Thu, 20 Apr 2023 23:18:43 GMT

This space is moving so fast my jaw drops every time I read something about it. For example this line: "... has discovered empirically that the capability for moral self-correction emerges at around 22B parameters. For more scientific details, the paper is here. [https://arxiv.org/pdf/2302.07459.pdf]"

Paper and discussion history. A mention is not an endorsement.

The Capacity for Moral Self-Correction in Large Language Models — HN submission

Thu, 16 Feb 2023 16:51:36 GMT

The Capacity for Moral Self-Correction in Large Language Models

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.