13 papers
Automated Numerical Stability Analysis of Deep Learning Operators
Xinye Chen
Finite-precision arithmetic unavoidably introduces numerical approximation errors. Numerical computations may use insufficient precision or an improper formulation, which leads to…
Floating-point autotuning with customized precisions
Xinye Chen, Thibault Hilaire, Fabienne Jézéquel
Reduced-precision arithmetic offers significant opportunities to improve performance, memory usage, and energy efficiency in numerical applications, provided that numerical accurac…
Using Laplace Transform To Optimize the Hallucination of Generation Models
Cheng Kang, Xinye Chen, Daniel Novak +1
To explore the feasibility of avoiding the confident error (or hallucination) of generation models (GMs), we formalise the system of GMs as a class of stochastic dynamical systems…
Estimating Condition Number with Graph Neural Networks
Erin Carson, Xinye Chen
In this paper, we propose a fast method for estimating the condition number of sparse matrices using graph neural networks (GNNs). For efficient deployment of GNNs, we introduce a…
QuantiPhy: A Quantitative Benchmark Evaluating Physical Reasoning Abilities of Vision-Language Models
Li Puyin, Tiange Xiang, Ella Mao +5
Understanding the physical world is essential for generalist AI agents. However, it remains unclear whether state-of-the-art vision perception models (e.g., large VLMs) can reason…
Mixed-Precision Conjugate Gradient Solvers with RL-Driven Precision Tuning
Xinye Chen
This paper presents a novel reinforcement learning (RL) framework for dynamically optimizing numerical precision in the preconditioned conjugate gradient (CG) method. By modeling p…
Modular Ideals and Factorizations of Affine Hyperplane Arrangements
Yanru Chen, Weikang Liang, Suijie Wang +1
Let be an affine hyperplane arrangement, let be its intersection semilattice, and let be its characteristic polynomial. We intro…
LLM-ABBA: Understanding time series via symbolic approximation
Xinye Chen, Erin Carson, Cheng Kang
The success of large language models (LLMs) for time series has been demonstrated in previous work. Utilizing a symbolic time series representation, one can efficiently bridge the…
QABBA: Symbolic Time-Series Compression via Integer-Quantized Aggregation
Erin Carson, Xinye Chen, Fei He +1
The expansion of time-series data from sensors and monitoring systems has made compact representations increasingly important. Such representations should retain signal structure w…
Mixed precision HODLR matrices
Erin Carson, Xinye Chen, Xiaobo Liu
Hierarchical matrix computations have attracted significant attention in the science and engineering community as exploiting data-sparse structures can significantly reduce the com…
Computing k-means in mixed precision
Erin Carson, Xinye Chen, Xiaobo Liu
Motivated by the increasing availability of low- and mixed-precision arithmetic on modern hardware, we develop mixed-precision variants of Lloyd's algorithm for k-means clustering.…
Quantized Embedding Vectors for Controllable Diffusion Language Models
Cheng Kang, Xinye Chen, Yong Hu +1
Improving the controllability, portability, and inference speed of diffusion language models (DLMs) is a key challenge in natural language generation. While recent research has sho…
Parallel Two-Stage Approach for Joint Symbolic Approximation of Time Series
Xinye Chen
As time-series applications grow larger, there is increasing demand for symbolic representations that are compact, accurate, and scalable across many signals and computing resource…