8 citations · 14 across the 6 of their papers we have counts for
7 papers
Inverse Reasoning Models for Efficient Post-Training and Verification from Answer-Only Data
Eric O. Korman
Post-training reasoning LLMs often requires making a tradeoff between costly supervision and compute inefficiency: supervised finetuning typically requires reasoning traces, which…
Atlas Based Representation and Metric Learning on Manifolds
Eric O. Korman
We explore the use of a topological manifold, represented as a collection of charts, as the target space of neural network based representation learning tasks. This is achieved by…
Autoencoding topology
Eric O. Korman
The problem of learning a manifold structure on a dataset is framed in terms of a generative model, to which we use ideas behind autoencoders (namely adversarial/Wasserstein autoen…
A hyperholomorphic line bundle on certain hyperkähler manifolds not admitting an -symmetry
Eric O. Korman
Generalizing work of Haydys and Hitchin, we prove the existence of a hyperholomorphic line bundle on certain hyperkähler manifolds that do not necessarily admit an action. As…
Characteristic classes of Higgs bundles and Reznikov's theorem
Eric O. Korman
We introduce Dolbeault cohomology valued characteristic classes of Higgs bundles over complex manifolds. Flat vector bundles have characteristic classes lying in odd degree de Rham…
Symplectic Dolbeault Operators on Kähler Manifolds
Eric O. Korman
For a Kähler Manifold , the "symplectic Dolbeault operators" are defined using the symplectic spinors and associated Dirac operators, in complete analogy to how the usual Dolbea…
Bilinear Forms and Fierz Identities for Real Spin Representations
Eric O. Korman, George Sparling
Given a real representation of the Clifford algebra corresponding to with metric of signature , we demonstrate the existence of two natural bilinear forms on the s…