I think you train Claude 3.7 to imitate the paraphrased scratchpad, but I’m a little unsure because you say “distill”. Just checking that Claude 3.7 still produces CoT (in the style of the paraphrase) after training, rather than being trained to perform the paraphrased-CoT reasoning in one step?
By distillation, I mean training to imitate. So in the distill-from-paraphrased setting, the only model involved at evaluation time is the base model fine-tuned on paraphrased scratchpads, and it generates an answer from beginning to end.
I think you train Claude 3.7 to imitate the paraphrased scratchpad, but I’m a little unsure because you say “distill”. Just checking that Claude 3.7 still produces CoT (in the style of the paraphrase) after training, rather than being trained to perform the paraphrased-CoT reasoning in one step?
By distillation, I mean training to imitate. So in the distill-from-paraphrased setting, the only model involved at evaluation time is the base model fine-tuned on paraphrased scratchpads, and it generates an answer from beginning to end.