Research

Quantitative finance

Hamiltonian Alpha Engine

Structured market dynamics for options research, connecting geometric forecasts to portfolio outcomes under costs, risk and causal evaluation.

Quantitative finance

Production Option-Surface Engine

A production-oriented option-surface system with a coherent price representation, analytic sensitivities and reproducible numerical validation.

NeurIPS 2026 Main Track–Poster

Machine Learning

Beyond Isotropy in JEPAs: Hamiltonian Geometry and Symplectic Prediction

JEPAs usually target an isotropic latent cloud; HamJEPA instead predicts between views with a symplectic map on phase space.

Quantitative finance

Arbitrage-Free Option Price Surfaces via Chebyshev Tensor Bases and a Hamiltonian Fog Post-Fit

We fit an arbitrage-carrying forward-discounted call surface directly in price space, enforce static no-arbitrage through sparse linear operators, and use a local fog layer where quotes are inconsistent.

Geometric Analysis

Rational-slope rigidity with identity holonomy through finite-amplitude resonant cells

Identity-holonomy minimizing foliations at every rational rotation force zero vertical force; ordinary all-slope foliations do not.

Geometric Analysis

Minimizing foliations: counterexamples and common-volume rigidity

All-slope minimizing foliations coexist with nonzero force. A common preserved volume yields a sharp slope-span criterion for rigidity.

Geometric Analysis

Variational rigidity for alternating symplectic billiards

A smooth convex body is totally integrable in its own gauge exactly when it is an origin-centred ellipse. Positive curvature follows from the dynamics.

Machine Learning

Prediction Is Not a Sufficient Memory State

Identical current predictions can hide different future updates. Sharp memory requirements reveal the evidence that cumulative learning and attention must preserve.