Re:Claude Fable 5.2 solves two open problems in the physics ...
by hgoel on Thursday November 19, 2026 @05:50PM
(#50567657)
Gigabytes of text seems perfectly shareable nowadays. Just as with most computational physics stuff of that scale, sharing code that can reproduce the result and roughly describing the computational requirements would be fine.
Re:Claude Fable 5.2 solves two open problems in the physics ...
by kulahan on Thursday November 19, 2026 @08:07PM
(#50568871)
LLMs are not deterministic, which to my dumb mind, seems to lessen the value of those inputs since they will not inherently get you down the same path, no?
Re:Claude Fable 5.2 solves two open problems in the physics ...
by inkysigma on Thursday November 19, 2026 @11:34PM
(#50570220)
I imagine the value here would be developing general strategies on how to use the tool to accomplish a task rather than reproducibility. You could probably make an LLM deterministic if need be but with significant performance penalty.
I wonder how this works out in other parts of physics codes. If you need a Monte Carlo integrator, is it common to guarantee a seed for reproducibility? How do these codes handle distributed systems for rng? Is it common to use a fully reproducible rng and GPU kernel for scientific compute?
Re:Claude Fable 5.2 solves two open problems in the physics ...
by bmenrigh on Thursday November 19, 2026 @09:11PM
(#50569314)
Oh it’s totally shareable. And I’d love to see it all shared.
But it isn’t “just a prompt”, and sharing it doesn’t make the results reproducible (since you don’t get control over the model’s “seed” from run to run).
Re:Claude Fable 5.2 solves two open problems in the physics ...
by JBorrow on Thursday November 19, 2026 @07:36PM
(#50568627)
It is common in physics and astronomy to share your code that takes millions of cpu hours to solve problems, and to share your resulting data that can be in the petabytes.
Re:Claude Fable 5.2 solves two open problems in the physics ...
by suddenlybananas on Friday November 20, 2026 @04:17AM
(#50572157)
If the human is involved, "where what the human is saying to guide the model", then the prompts are even more important to include. Is this a case of "solve an open problem in theoretical physics, make no mistakes", or a case where a human had to interact frequently with the model?
For any of the big things I've worked on, it isn't one "prompt", it's a goal that keeps getting updated and a bunch of back-and-forth steering of the model and sub agents, where what the human is saying to guide the model is heavily context dependent on what the model is currently doing or saying. Much of that guidance will ultimately be useless guidance on dead-end explorations. Also, the model will likely have written huge programs to solve things symbolically or simulate things numerically, and both the programs that were written, and the output of those programs are critical to understanding how the model came up with the things that led to a solution.
Now, I'd be all for publishing the full transcript, but the transcript is likely to be massive. Gigabytes of text, and reproducing the local computation (often sympy or numpy) may be hundreds or thousands of CPU-hours.