arXiv subjects

Mark Landry

Publications and source records attributed to Mark Landry.

6 recordsLinked to original sources

arXiv2026

TabH2O: A Unified Foundation Model for Tabular Prediction

We present TabH2O, a foundation model for tabular data that performs classification and regression in a single forward pass via in-context learning. TabH2O builds on the TabICL architecture with several key modifications: (1) unified training, a single model handles both classification and regression via a dual-head architecture, eliminating the need for separate models and reducing total pretraining cost; (2) single-stage pretraining, training stability improvements (bounded scalable softmax, inter-stage normalization, learnable residual scaling, logit soft-capping) eliminate the need for multi-stage curriculum learning, enabling training with full-length sequences from the start; and (3) noise-aware pretraining, synthetic datasets include explicit noise dimensions to teach the model robustness to irrelevant features. We evaluate TabH2O v1.1 (29.2M parameters) on the TALENT benchmark (300 datasets), where it achieves an average rank of 2.54 out of 6 evaluated methods, outperforming tuned CatBoost (4.00), H2O AutoML (4.28), LightGBM (4.98), and TabPFN v2.6 (2.80), and competitive compared to TabICL v2 (2.13), while placing in the top-3 on 80% of the testing datasets across classification and regression tasks. We also achieve state-of-the-art results on TabArena. Moreover, our single model is well-suited to tabular datasets up to roughly 500,000 rows and 100 features.

cs.LG↗

arXiv2024

H2OVL-Mississippi Vision Language Models Technical Report

Smaller vision-language models (VLMs) are becoming increasingly important for privacy-focused, on-device applications due to their ability to run efficiently on consumer hardware for processing enterprise commercial documents and images. These models require strong language understanding and visual capabilities to enhance human-machine interaction. To address this need, we present H2OVL-Mississippi, a pair of small VLMs trained on 37 million image-text pairs using 240 hours of compute on 8 x H100 GPUs. H2OVL-Mississippi-0.8B is a tiny model with 0.8 billion parameters that specializes in text recognition, achieving state of the art performance on the Text Recognition portion of OCRBench and surpassing much larger models in this area. Additionally, we are releasing H2OVL-Mississippi-2B, a 2 billion parameter model for general use cases, exhibiting highly competitive metrics across various academic benchmarks. Both models build upon our prior work with H2O-Danube language models, extending their capabilities into the visual domain. We release them under the Apache 2.0 license, making VLMs accessible to everyone, democratizing document AI and visual LLMs.

cs.CV↗

arXiv2023

h2oGPT: Democratizing Large Language Models

Applications built on top of Large Language Models (LLMs) such as GPT-4 represent a revolution in AI due to their human-level capabilities in natural language processing. However, they also pose many significant risks such as the presence of biased, private, or harmful text, and the unauthorized inclusion of copyrighted material. We introduce h2oGPT, a suite of open-source code repositories for the creation and use of LLMs based on Generative Pretrained Transformers (GPTs). The goal of this project is to create the world's best truly open-source alternative to closed-source approaches. In collaboration with and as part of the incredible and unstoppable open-source community, we open-source several fine-tuned h2oGPT models from 7 to 40 Billion parameters, ready for commercial use under fully permissive Apache 2.0 licenses. Included in our release is 100\% private document search using natural language. Open-source language models help boost AI development and make it more accessible and trustworthy. They lower entry hurdles, allowing people and groups to tailor these models to their needs. This openness increases innovation, transparency, and fairness. An open-source strategy is needed to share AI benefits fairly, and H2O.ai will continue to democratize AI and LLMs.

cs.CL↗

arXiv2020

Interacting particle systems with type D symmetry and duality

We construct a two-class asymmetric interacting particle system with $U_q(so_6)$ or $U_q(so_8)$ symmetry, in which up to two particles may occupy a site if the two particles have different class. The particles exhibit a drift, but there is no preference given between first-class and second-class particles. The quantum group symmetry leads to reversible measures and a self-duality for the particle system. Additionally, a new method is developed to construct a symmetric interacting particle system from the Casimir element of $so_{2n}$.

math.PR↗

arXiv2020

The Multiple Points of Fractional Brownian Motion

Nils Tongring (1987) proved sufficient conditions for a compact set to contain $k$-tuple points of a Brownian motion. In this paper, we extend these findings to the fractional Brownian motion. Using the property of strong local nondeterminism, we show that if $B$ is a fractional Brownian motion in $\mathbb{R}^d$ with Hurst index $H$ such that $Hd=1$, and $E$ is a fixed, nonempty compact set in $\mathbb{R}^d$ with positive capacity with respect to the function $\phi(s) = (\log_+(1/s))^k$, then $E$ contains $k$-tuple points with positive probability. For the $Hd > 1$ case, the same result holds with the function replaced by $\phi(s) = s^{-k(d-1/H)}$.

math.PR↗

arXiv2019

Landscape Theory for Schr\"odinger Operators with General Hopping Terms on a Finite Lattice

Findings by M. L. Lyra, S. Mayboroda and M. Filoche relate invertibility and positivity of a class of discrete Schr\"odinger matrices with the existence of the "Landscape Function", which provides an upper bound on all eigenvectors simultaneously. Their argument is based on the variational principles. We consider an alternative method of proving these results, based on the power series expansion, and demonstrate that it naturally extends the original findings to the case of long range operators. The method of proof by power series expansion can also be employed in other scenarios, such as higher dimensional lattices.

math-ph↗