From c3c4539553c2fd957e148a94c6d5522ba9b5fda6 Mon Sep 17 00:00:00 2001 From: ChrisRackauckas Date: Thu, 16 Oct 2025 00:09:50 -0400 Subject: [PATCH] Add Chris Rackauckas publications from 2023-2025 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Added five recent publications with Chris Rackauckas as co-author: 1. "Efficient Symbolic Computation via Hash Consing" (arXiv:2509.20534, 2025) - Presents integration of hash consing into JuliaSymbolics - Authors: Zhu, Sabharwal, Tan, Ma, Edelman, Rackauckas 2. "The Future of Artificial Intelligence and the Mathematical and Physical Sciences" (arXiv:2509.02661, 2025) - Community paper from NSF Workshop on AI+MPS - 100+ contributors including Rackauckas 3. "Physics-Constrained Flow Matching" (arXiv:2506.04171, 2025) - Zero-shot inference framework for generative models with constraints - Authors: Utkarsh, Cai, Edelman, Gomez-Bombarelli, Rackauckas 4. "Semi-Explicit Neural DAEs" (arXiv:2505.20515, 2025) - Method for enforcing algebraic constraints in neural differential equations - Authors: Pal, Edelman, Rackauckas 5. "Robust Parameter Estimation for Rational ODEs" (arXiv:2303.02159, 2023) - Published in Applied Mathematics and Computation, 2026 - Authors: Bassik, Berman, Go, Hong, Ilmer, Ovchinnikov, Rackauckas, Soto, Yap These additions expand the bibliography with recent work in scientific machine learning, symbolic computation, neural differential equations, and generative models with physical constraints. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude --- _assets/julialab.bib | 58 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+) diff --git a/_assets/julialab.bib b/_assets/julialab.bib index 0f4ddb0..c09c4fb 100644 --- a/_assets/julialab.bib +++ b/_assets/julialab.bib @@ -10,6 +10,64 @@ % url={url to the journal/conference page of your article}, % +@misc{zhu2025hashconsing, + title={Efficient Symbolic Computation via Hash Consing}, + author={Zhu, Bowen and Sabharwal, Aayush and Tan, Songchen and Ma, Yingbo and Edelman, Alan and Rackauckas, Christopher}, + year={2025}, + eprint={2509.20534}, + archivePrefix={arXiv}, + primaryClass={cs.PL}, + url={https://arxiv.org/abs/2509.20534}, + abstract={Symbolic computation systems suffer from memory inefficiencies due to redundant storage of structurally identical subexpressions. This work presents the first integration of hash consing into JuliaSymbolics, employing a global weak-reference hash table that canonicalizes expressions and eliminates duplication, reducing memory consumption and accelerating operations such as differentiation, simplification, and code generation.} +} + +@misc{ferguson2025aimps, + title={The Future of Artificial Intelligence and the Mathematical and Physical Sciences (AI+MPS)}, + author={Ferguson, Andrew and LaFleur, Marisa and Ruthotto, Lars and Thaler, Jesse and Ting, Yuan-Sen and Tiwary, Pratyush and Villar, Soledad and others}, + year={2025}, + eprint={2509.02661}, + archivePrefix={arXiv}, + primaryClass={cs.AI}, + url={https://arxiv.org/abs/2509.02661}, + note={Community Paper from NSF Future of AI+MPS Workshop, Cambridge, Massachusetts, March 24-26, 2025. Christopher Rackauckas (JuliaHub, Pumas-AI, MIT) listed as contributing author.}, + abstract={Community paper from NSF Workshop examining how mathematical and physical science domains can leverage AI while contributing to its development. Proposes strategic priorities for bidirectional AI+MPS research, interdisciplinary community building, and workforce development.} +} + +@misc{utkarsh2025pcfm, + title={Physics-Constrained Flow Matching: Sampling Generative Models with Hard Constraints}, + author={Utkarsh, Utkarsh and Cai, Pengfei and Edelman, Alan and Gomez-Bombarelli, Rafael and Rackauckas, Christopher Vincent}, + year={2025}, + eprint={2506.04171}, + archivePrefix={arXiv}, + primaryClass={cs.LG}, + url={https://arxiv.org/abs/2506.04171}, + abstract={Deep generative models have recently been applied to physical systems governed by partial differential equations (PDEs), offering scalable simulation and uncertainty-aware inference. However, enforcing physical constraints, such as conservation laws and physical consistencies, remains challenging. This work proposes Physics-Constrained Flow Matching (PCFM), a zero-shot inference framework that enforces arbitrary nonlinear constraints in pretrained flow-based generative models while ensuring exact constraint satisfaction.} +} + +@misc{pal2025neuralDAEs, + title={Semi-Explicit Neural DAEs: Learning Long-Horizon Dynamical Systems with Algebraic Constraints}, + author={Pal, Avik and Edelman, Alan and Rackauckas, Christopher}, + year={2025}, + eprint={2505.20515}, + archivePrefix={arXiv}, + primaryClass={cs.LG}, + url={https://arxiv.org/abs/2505.20515}, + abstract={A method for enforcing algebraic constraints in neural differential equations by projecting ODE steps onto constraint manifolds, achieving constraint violation errors below 10^{-10} across benchmark problems.} +} + +@misc{bassik2023robust, + title={Robust Parameter Estimation for Rational Ordinary Differential Equations}, + author={Bassik, Oren and Berman, Yosef and Go, Soo and Hong, Hoon and Ilmer, Ilia and Ovchinnikov, Alexey and Rackauckas, Chris and Soto, Pedro and Yap, Chee}, + year={2023}, + eprint={2303.02159}, + archivePrefix={arXiv}, + primaryClass={cs.MS}, + url={https://arxiv.org/abs/2303.02159}, + doi={10.1016/j.amc.2025.129638}, + note={Published in Applied Mathematics and Computation, Vol. 509, 15 January 2026}, + abstract={A novel approach for estimating parameters in rational ODE models from time series data using differential algebra, rational function interpolation, and multivariate polynomial system solving, avoiding dependence on initial guesses and search intervals.} +} + @misc{ringoot2025gpuresidentmemoryawarealgorithmaccelerating, title={A GPU-resident Memory-Aware Algorithm for Accelerating Bidiagonalization of Banded Matrices}, author={Evelyne Ringoot and Rabab Alomairy and Alan Edelman},