traceoid.ai is scaling energy-based models (EBMs) by revisiting the math that underlies machine learning. Here is what Yann LeCun had to say about EBMs. Our interpretation of EBMs will provide: interpretability, composability 1, faster & cheaper training and inference etc.
Our approach will mark a significant step forward towards achieving AGI.
Our thesis is that there is 100 years of physics and math research that has gone unnoticed by the CS/ML communities and we intend to rectify that.
Join our Discord channel.
We are building a language and a compiler tailored towards optimization and machine learning, in particular EBMs. Our approach leverages advanced approaches from math and physics to achieve the desirable properties above.
Please reach out if you have background in one or ideally more of the following (and related) fields (or are excited to learn):
- integrable systems
- algebraic topology (in particular Hopf algebras, tensor categories)
- category theory
- functional programming
- compilers
- statistical mechanics
- computational physics
- quantum groups
- matroid theory
- harmonic analysis
- theorem provers
- computer algebra systems
- numerical computation
- algebraic geometry
Ideal candidate would understand how one goes about writing high perf code as well as be able to explain the concept of a Hamiltonian. You will be joining a team with several PhDs and while advanced degrees are not a hard requirement, you are expected to possess deep knowledge of your field.
If you have experience with any of the above (or related), but have no machine learning experience, and are consider applying, please do not hesitate and reach out.
Our company operates at the intersection of research & development, and as a part of this job, you will be reading papers and implementing some very cool research from the above fields.
If the job sounds interesting, please do not hesitate and contact us at adam+careers@traceoid.ai or over Discord.
All positions are fully remote.
We recently finished a VC funding round and we are currently not raising additional funding. However, if you might be interested in investing in the future, please reach out now, either over email at adam+vc@traceoid.ai or over Discord.
Footnotes
-
Creating larger models from smaller models. ↩