search

spcl / graph-of-thoughts

Official Implementation of "Graph of Thoughts: Solving Elaborate Problems with Large Language Models"
https://arxiv.org/pdf/2308.09687.pdf
Other
1.97k stars 141 forks source link

thanks for the cite, much appreciated #3

Closed qrdlgit closed 9 months ago

qrdlgit commented 10 months ago

It's always great to see a lack of ego and respect for others in academia.

nblach commented 9 months ago

Hi @qrdlgit,

We checked your profile and saw that you have a repository with the same name, but fail to see how our works relate beyond the common base (ToT).

You are applying ToT-BFS with a branching factor of 3 and customisable depth to improve an ML program, whereby the scoring function reports the achieved r2 score.

This does not intersect with our work, as we are providing a general framework that improves upon ToT with novel ideas and does not cover the automatic improvement of an ML program.

Nevertheless, we'll be happy to acknowledge your work in our paper - please send us a reference to be included (is there a corresponding paper/draft, or shall we just cite your repo?).

qrdlgit commented 9 months ago

In general, citing git repos is pretty standard in papers these days.

The approach I take is almost exactly what your paper does, except that I published the repo 4 months before your paper.

Honestly, I'm less concerned about the lack of cite to my stuff, but rather the fact that a lot of papers only cite if there is latex behind it. A lot of ideas I see in arxiv these days are derivative of stuff already in the ecosystem, except missing a reference.

There needs to be a standard / section where folks reference these types of things. I don't necessarily think a discussion of results is required - given that the results are often informal like mine - but some acknowledgement that a survey was performed at a minimum should be required. Just something like -

Other informal efforts were surveyed - [1] [2] [3] [4]

would be enough.

The fact that you folks didn't find mine (literally named 'graph of thoughts') makes me wonder what else you missed.