STaR: Bootstrapping Reasoning With Reasoning · HackerPapers discussions

STaR: Bootstrapping Reasoning With Reasoning — mentioned by viraptor

Sun, 22 Dec 2024 22:17:45 GMT

This is a good read for some examples https://arxiv.org/abs/2203.14465 > This technique, the "Self-Taught Reasoner" (STaR), relies on a simple loop: generate rationales to answer many questions, prompted with a few rationale examples; if the generated answers are wrong, try again to generate a rationale given the correct answer; fine-tune on all the rationales that ultimately yielded correct answers; repeat. We show that STaR significantly improves performance on multiple datasets compared to a model fine-tuned to directly predict final answers But there are a few others. In general good data…

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by viraptor

Mon, 14 Oct 2024 05:51:21 GMT

It's been already proven possible https://arxiv.org/abs/2203.14465

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by imtringued

Sun, 01 Sep 2024 09:02:53 GMT

It is in reference to this: https://arxiv.org/abs/2203.14465 https://arxiv.org/abs/2403.09629

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by krackers

Sat, 31 Aug 2024 20:50:41 GMT

According to that recent The Information report, Orion is supposed to be just a regular LLM except trained with synthetic data generated via Strawberry. Anthropic et al. have also been working on ways to generate synthetic data (as seen in the success of Sonnet 3.5) so I don't really know if that's going to be a big lead. And of course the ever-hyped strawberry is supposed to be some sort of tree-of-thought type thing I think, or maybe it's related to https://arxiv.org/abs/2203.14465. Either way, nothing so far has come out that it's a completely novel training technique or architecture, just…

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by seanhunter

Wed, 14 Aug 2024 10:21:44 GMT

My understanding is this used to be the case[1] but isn't really true any longer due to things like the "star" method for model training[2]. Empirically it absolutely (circa GPT3) used to be the case that if you prompted with "Explain all your reasoning step by step and then give the answer at the end" or similar it would give you a better answer for a complex question than if you said "Just give me the answer and nothing else" or similar, or asked for the answer first, and then circa gpt-4 answers started getting much longer even if you asked the model to be concise. That doesn't seem to be…

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by viraptor

Wed, 24 Jul 2024 14:08:33 GMT

- Throwing out the bad parts, or updating the good parts with better responses. (There's a few training programming data sets which are not great) - Self improvement without new data, like STaR https://arxiv.org/abs/2203.14465 - architecture changes to support better generalisation - humans learn a lot without consuming as much content.

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by peter_d_sherman

Sat, 02 Dec 2023 05:21:34 GMT

Specific points of time in video: "What is Q<star>?": https://www.youtube.com/watch?v=Z6E41eXStsU&t=329s "LLM's [AI's] today still don't reason very well": https://www.youtube.com/watch?v=Z6E41eXStsU&t=405s Peter Liu, GSM8K/STaR tweet: https://www.youtube.com/watch?v=Z6E41eXStsU&t=465s "STaR: Bootstrapping Reasoning With Reasoning" https://arxiv.org/abs/2203.14465 (arxiv: cs > arXiv:2203.14465, Computer Science > Machine Learning, Eric Zelikman, Yuhuai Wu, Jesse Mu, Noah D. Goodman, 2022) >"We propose a technique to iteratively leverage a small number of rationale examples and a large dataset…

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

STaR: Bootstrapping Reasoning With Reasoning — mentioned by gwern

Wed, 30 Mar 2022 02:34:08 GMT

Yeah, but I didn't bring it up because I wasn't sure how much is really the model choosing and how much is the human workflow: they emphasize the interactive part heavily. Anyway, today another great paper dropped on self-distillation: "STaR: Bootstrapping Reasoning With Reasoning" https://arxiv.org/abs/2203.14465 , Zelikman et al 2022.

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