Base Models Beat Aligned Models at Randomness and Creativity · HackerPapers discussions

Base Models Beat Aligned Models at Randomness and Creativity — mentioned by stratos123

Thu, 15 Oct 2026 17:12:56 GMT

> Interesting. I'd love to see a paper on this. The first time I personally heard of RLHF ruining some model qualities was actually in the GPT-4 paper (march 2023); figure 8 there shows that the base model has near-perfect calibration on multiple-choice questions (!) and this stops being true for the RLHFed model: https://arxiv.org/abs/2303.08774v6 I think this is the same sort of problem - that the miscalibration of RLHFed models also explains their lack of diversity. In the training dataset, for a prompt like "you're a helpful assistant, yadda yadda, write a site", there's a large range of…

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

Base Models Beat Aligned Models at Randomness and Creativity — mentioned by deepsquirrelnet

Wed, 25 Feb 2026 19:47:12 GMT

Ask an llm to pick a random number from 1-10. My money is on 7. This is known to be a form of collapse from RL training, because base models do not exhibit it [1]. 1. https://arxiv.org/abs/2505.00047

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

Base Models Beat Aligned Models at Randomness and Creativity — HN submission

Tue, 13 May 2025 07:42:48 GMT

Base Models Beat Aligned Models at Randomness and Creativity

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

Base Models Beat Aligned Models at Randomness and Creativity — HN submission

Wed, 07 May 2025 17:43:00 GMT

Base Models Beat Aligned Models at Randomness and Creativity

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