arXiv subjects

Ryan Krueger

Publications and source records attributed to Ryan Krueger.

4 recordsLinked to original sources

arXiv2025

An AI system to help scientists write expert-level empirical software

The cycle of scientific discovery is frequently bottlenecked by the slow, manual creation of software to support computational experiments\cite{hannay2009how}. To address this, we present Empirical Research Assistance (ERA), an AI system that creates expert-level scientific software whose goal is to maximize a quality metric. The system uses a Large Language Model (LLM) and Tree Search (TS)\cite{silver2016mastering} to systematically improve the quality metric and intelligently navigate the large space of possible solutions. ERA achieves expert-level results when it explores and integrates complex research ideas from external sources. The effectiveness of tree search is demonstrated across a diverse range of tasks. In bioinformatics, ERA discovered 40 novel methods for single-cell data analysis that outperformed the top human-developed methods on a public leaderboard. In epidemiology, ERA generated 14 models that outperformed the CDC ensemble and all other individual models for forecasting COVID-19 hospitalizations. ERA also produced expert-level software for geospatial analysis, neural activity prediction in zebrafish, and numerical solution of integrals, and a novel rule-based construction for time series forecasting. By devising and implementing novel solutions to diverse tasks, ERA represents a significant step towards accelerating scientific progress.

cs.AI↗

arXiv2023

Tuning Colloidal Reactions

The precise control of complex reactions is critical for biological processes yet our inability to design for specific outcomes limits the development of synthetic analogues. Here, we leverage differentiable simulators to design nontrivial reaction pathways in colloidal assemblies. By optimizing over external structures, we achieve controlled disassembly and particle release from colloidal shells. Lastly, we characterize the role of configurational entropy in the structure via both forward calculations and optimization, inspiring new parameterizations of designed colloidal reactions.

cond-mat.soft↗

arXiv2022

Vanishing 2-Qubit Gates with Non-Simplification ZX-Rules

Traditional quantum circuit optimization is performed directly at the circuit level. Alternatively, a quantum circuit can be translated to a ZX-diagram which can be simplified using the rules of the ZX-calculus, after which a simplified circuit can be extracted. However, the best-known extraction procedures can drastically increase the number of 2-qubit gates. In this work, we take advantage of the fact that local changes in a ZX-diagram can drastically affect the complexity of the extracted circuit. We use a pair of congruences (i.e., non-simplification rewrite rules) based on the graph-theoretic notions of local complementation and pivoting to generate local variants of a simplified ZX-diagram. We explore the space of equivalent ZX-diagrams generated by these congruences using simulated annealing and genetic algorithms to obtain a simplified circuit with fewer 2-qubit gates. On randomly generated circuits, our method can outperform state-of-the-art optimization techniques for low-qubit (<10) circuits. On a set of previously reported benchmark circuits with <=14 qubits, our method outperforms off-the-shelf methods in 87% of cases, consistently reducing overall circuit complexity by an additional ~15-30% and eliminating up to 46% of 2-qubit gates. These preliminary results serve as a proof-of-concept for a new circuit optimization strategy in the ZX-calculus.

quant-ph↗

arXiv2021

Automatically Building Diagrams for Olympiad Geometry Problems

We present a method for automatically building diagrams for olympiad-level geometry problems and implement our approach in a new open-source software tool, the Geometry Model Builder (GMB). Central to our method is a new domain-specific language, the Geometry Model-Building Language (GMBL), for specifying geometry problems along with additional metadata useful for building diagrams. A GMBL program specifies (1) how to parameterize geometric objects (or sets of geometric objects) and initialize these parameterized quantities, (2) which quantities to compute directly from other quantities, and (3) additional constraints to accumulate into a (differentiable) loss function. A GMBL program induces a (usually) tractable numerical optimization problem whose solutions correspond to diagrams of the original problem statement, and that we can solve reliably using gradient descent. Of the 39 geometry problems since 2000 appearing in the International Mathematical Olympiad, 36 can be expressed in our logic and our system can produce diagrams for 94% of them on average. To the best of our knowledge, our method is the first in automated geometry diagram construction to generate models for such complex problems.

cs.CG↗