For those who are frustrated with the lack of a fleshed out limitations section, it seems well covered under the peer review which is in the link above or directly below [0]
To me the paper is still very interesting, but concerns about computationally intractability and the hardness of approximation questions made me dig deeper.
> Specifically, our model aims to bridge biological circuits and computations of uncertainty in a tractable manner.
> To address this, we have carefully reframed our claims throughout the manuscript to emphasize that the model is a hypothesis generator rather than a definitive representation of biological circuitry.
Under the "All models are wrong, some are useful" I have no doubt this will be useful to some. But I will admit that their claims in the response to Reviewer Comment 4.5 that they "emphasize that the model is a hypothesis generator" doesn't match the published paper IMHO; and that negatively impacted my view of the claims in an admittedly probably unfair manner.
To me the paper is still very interesting, but concerns about computationally intractability and the hardness of approximation questions made me dig deeper.
> Specifically, our model aims to bridge biological circuits and computations of uncertainty in a tractable manner.
> To address this, we have carefully reframed our claims throughout the manuscript to emphasize that the model is a hypothesis generator rather than a definitive representation of biological circuitry.
Under the "All models are wrong, some are useful" I have no doubt this will be useful to some. But I will admit that their claims in the response to Reviewer Comment 4.5 that they "emphasize that the model is a hypothesis generator" doesn't match the published paper IMHO; and that negatively impacted my view of the claims in an admittedly probably unfair manner.
[0] https://static-content.springer.com/esm/art%3A10.1038%2Fs414...