Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping · HackerPapers discussions

Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping — mentioned by hislaziness

Wed, 23 Oct 2024 05:03:31 GMT

As I understand, the LLM uses the techniques of searchformer - https://arxiv.org/abs/2402.14083. To do "slow thinking" doing a A* search using a transofrmer.

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

Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping — mentioned by throwaway4aday

Tue, 10 Sep 2024 21:28:57 GMT

While I agree that we don't want to extrapolate too much I disagree that this type of exploration may not benefit from more compute. We won't know until we try and since we have what seems to be a very generalizable architecture it makes sense to take the brute force approach of creating models of that data by scaling the amount of data and the amount of compute we dedicate to it. If it turns out not to work then we've learned something. As it turns out, Logic and Algorithms has seen some early success using Transformers (Searchformer) https://arxiv.org/abs/2402.14083

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

Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping — HN submission

Fri, 23 Feb 2024 22:34:11 GMT

Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping

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

Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping — HN submission

Fri, 23 Feb 2024 11:53:25 GMT

Beyond A*: Better Planning with Transformers

313 points; 120 discussion replies on HN. Paper and source history. A submission is not an endorsement.