Will a matmul algorithm better than O(n^{2.371552}) [Willians et al. 2023] be published before the end of 2025?
TechnologyScienceComputational Complexity Theory
Basic
26
Ṁ1.0k
2025
71%
chance

To date the best announced bound on the asymptotic complexity for matrix multiplication is 2.371552, published as a preprint [1].

Before the end of 2025, will a better bound appear in PEER REVIEWED venue?

Preprints will not be accepted as sufficient evidence, since there are perverse market incentives and verification issues for resolution.

[1] https://cs.paperswithcode.com/paper/new-bounds-for-matrix-multiplication-from

Papers with Code - New Bounds for Matrix Multiplication: from Alpha to Omega
No code available yet.
Get Ṁ600 play money

Related questions

9 Comments
22 Holders
41 Trades
Sort by:

i believe there is one now? new bound: 2.371339

https://arxiv.org/pdf/2404.16349

What if the algorithm is accepted to a peer-reviewed conference by EOY but the proceedings haven’t been published yet? E.g. SODA is in January.

Is it by the end of 2024 as in the description or 2025 as in the title?

predicts YES

@AntoineTilloy 2025, description was a typo, thanks for the catch. Most traders will have looked at title and resolution time, not description.

@Widden sure, thanks, was just asking to be sure

Researchers love to find "better" algorithms for matrix multiplication.

predicts YES

They cannot be stopped.

@jskf Market will still resolve yes if the bound is only valid for N>10e100e100, on every 2nd Tuesday of the month, while standing on one leg

predicts YES

Yes I know 😭

Related questions