Dynamic Bayesian Optimization Framework for Instruction Tuning in Partial Differential Equation Discovery
Not on the map. Why?
Abstract (arXiv)
Large Language Models (LLMs) show promise for equation discovery, yet their outputs are highly sensitive to prompt phrasing, a phenomenon we term instruction brittleness. Static prompts cannot adapt to the evolving state of a multi-step generation process, causing models to plateau at suboptimal solutions. To address this, we propose NeuroSymBO, which reframes prompt engineering as a sequential decision problem. Our method maintains a discrete library of reasoning strategies and uses Bayesian Optimization to select the optimal instruction at each step based on numerical feedback. Experiments on PDE discovery benchmarks show that adaptive instruction selection significantly outperforms fixed prompts, achieving higher recovery rates with more parsimonious solutions.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 30 of its 65 references that are in the map. First posted on arXiv 31 Dec 2025.
Builds on
30 of 65 references.
References · 30
- 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
- 2021Evaluating Large Language Models Trained on Codecited by 8,320
- 2022Large Language Models Are Human-Level Prompt Engineerscited by 1,045
- 2023Large Language Models as Optimizerscited by 864
- 2023Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formattingcited by 817
- 2023Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolutioncited by 551
- 2019AI Feynman: a Physics-Inspired Method for Symbolic Regressioncited by 548
- 2020Discovering Symbolic Models from Deep Learning with Inductive Biasescited by 344
- 2023State of What Art? A Call for Multi-Prompt LLM Evaluationcited by 310
- 2019Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradientscited by 309
- 2021Contemporary Symbolic Regression Methods and their Relative Performancecited by 215
- 2022End-to-end symbolic regression with transformerscited by 203
18 more
- 2021Neural Symbolic Regression that Scalescited by 192
- 2020Weak SINDy For Partial Differential Equationscited by 155
- 2020AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularitycited by 144
- 2021Ensemble-SINDy: Robust sparse model discovery in the low-data, high-noise limit, with active learning and controlcited by 137
- 2024LLM-SR: Scientific Equation Discovery via Programming with Large Language Modelscited by 133
- 2024ChemLLM: A Chemical Large Language Modelcited by 106
- 2021Symbolic Regression via Neural-Guided Genetic Programming Population Seedingcited by 101
- 2024LLM and Simulation as Bilevel Optimizers: A New Paradigm to Advance Physical Scientific Discoverycited by 86
- 2024Symbolic Regression with a Learned Concept Librarycited by 83
- 2018Bayesian Optimization of Combinatorial Structurescited by 64
- 2022SymFormer: End-to-end symbolic regression using transformer-based architecturecited by 59
- 2024Prompt Optimization with EASE? Efficient Ordering-aware Automated Selection of Exemplarscited by 37
- 2020A Survey of Deep Learning for Scientific Discoverycited by 36
- 2024In-Context Symbolic Regression: Leveraging Large Language Models for Function Discoverycited by 31
- 2024LLM4ED: Large Language Models for Automatic Equation Discoverycited by 14
- 2023Equation Discovery with Bayesian Spike-and-Slab Priors and Efficient Kernelscited by 5
- 2023A Bayesian approach for prompt optimization in pre-trained language modelscited by 4
- 2025(Exhaustive) Symbolic Regression and model selection by minimum description lengthcited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.