Automating Parent Selection Configuration in Genetic Programming with Agentic AI

Jose Guadalupe Hernandez; Jui-Hsuan Chang; Anil Kumar Saini; Xi Li; Jason H. Moore

cs.NEAug 17, 2026

Automating Parent Selection Configuration in Genetic Programming with Agentic AI

Authors: Jose Guadalupe Hernandez, Jui-Hsuan Chang, Anil Kumar Saini, Xi Li, Jason H. Moore

Organizations: Department of Computational Biomedicine, Cedars-Sinai Medical Center, West Hollywood, CA, 90069, USA.

Abstract

We investigate whether agentic artificial intelligence can automate parts of the process of designing genetic programming systems by introducing an agentic framework that identifies and implements parent selection algorithms using large language model (LLM) reasoning and retrieval-augmented generation. Using symbolic regression as a test bed, we first conduct an ablation study across four LLM types to evaluate the effects of agentic reasoning and retrieval on generated algorithm categories, validity, implementation similarity, and downstream performance. Results show that these components substantially influence the types of algorithms generated, but their downstream performance largely depends on the underlying LLM. The strongest configuration, the full agentic setup with 5 mini (5 mini--AR), consistently generated established $ε$ -lexicase implementations while maintaining competitive downstream performance. We then benchmark this configuration against fixed implementations of tournament selection and semi-dynamic MAD $ε$ -lexicase. Across six symbolic regression problems, 5 mini--AR performed similarly to $ε$ -lexicase while generally outperforming tournament selection. These findings demonstrate the potential of agentic AI to translate domain knowledge into generating executable components, providing a step toward automated configuration and design of evolutionary systems.

Figures & tables

Fig. 1: Overview of the proposed agentic framework for identifying and implementing parent selection algorithms: (a) research literature is collected into a curated corpus, (b) documents are loaded and partitioned into text chunks, (c) chunks are embedded and stored in a vector database, (d) the LLM-based agent iteratively reasons and may retrieve relevant literature to inform algorithm selection and implementation, and (e) the selected method, rationale, generated code, and associated metadata are recorded for subsequent analysis.

Parent Selection Method	NDS	RDS	IDS
Tournament	[ 38 , 47 , 48 , 40 , 45 , 52 , 55 , 56 ]	[ 52 , 48 ]	[ 52 ]
Fitness Proportionate	[ 38 , 52 ]	[ 52 ]	[ 52 ]
Lexicase	[ 47 , 40 , 45 , 52 , 51 , 55 ]	[ 52 , 55 , 51 ]	[ 52 ]
$\epsilon$ -Lexicase	[ 47 , 48 , 45 , 55 , 56 ]	[ 48 ]	-
$\epsilon$ -Plexicase	[ 55 , 48 ]	[ 48 ]	-
Batch Tournament	[ 48 , 56 ]	[ 48 ]	-

TABLE I: Research papers included in the RAG corpus and their coverage of parent selection methods under no down-sampling (NDS), Random Down-Sampling (RDS), and Informed Down-Sampling (IDS).

Task	Features	Samples
Airfoil	5	1503
Concrete	8	1030
Energy Cooling	8	768
Energy Heating	8	768
Housing	13	506
Yacht	6	308

TABLE II: Number of features and samples for all considered datasets.

Model	Agent	$\epsilon$ -Lexicase	Lexicase	Tournament	Unknown
Type	Config	Selection	Selection	Selection	Selection
oss 20b	LLM	0 (0.0%)	0 (0.0%)	126 (100.0%)	0 (0.0%)
	ANR	0 (0.0%)	0 (0.0%)	126 (100.0%)	0 (0.0%)
	AR	56 (44.4%)	36 (28.6%)	29 (23.0%)	5 (4.0%)
4o	LLM	0 (0.0%)	0 (0.0%)	126 (100.0%)	0 (0.0%)
	ANR	0 (0.0%)	0 (0.0%)	126 (100.0%)	0 (0.0%)

TABLE III: Counts (percentages) of the parent selection algorithm families generated by each LLM and agent configuration. Each LLM–configuration pairing generated 126 algorithms. Boldface indicates the dominant family for each LLM–configuration pairing.

Model	Agent	$\epsilon$ -Lexicase	Static	Semi-Dynamic	Dynamic
Type	Config	Total	Count	Count	Count
oss 20b	AR	56	0 (0.0%)	17 (30.4%)	39 (69.6%)
4o	AR	96	0 (0.0%)	96 (100.0%)	0 (0.0%)
4.1 mini	AR	124	0 (0.0%)	107 (86.3%)	17 (13.7%)
5 mini	LLM	108	5 (4.6%)	91 (84.3%)	12 (11.1%)
	ANR	92	10 (10.9%)	56 (60.9%)	26 (28.3%)

TABLE IV: Counts (percentages) of the

\epsilon

-lexicase implementation variants among the algorithms classified as

\epsilon

-lexicase in Table III . Percentages are computed relative to the total number of

\epsilon

-lexicase algorithms generated for each LLM–configuration pairing. Only pairings that generated at least one

\epsilon

-lexicase algorithm are included.

Model	Agent	$\epsilon$ -lex	MAD	$\epsilon\geq 0.1$	$\epsilon<0.1$	Other
Type	Config	Total	Count	Count	Count	Count
oss 20b	AR	56	0 (0.0%)	0 (0.0%)	9 (16.1%)	47 (83.9%)
4o	AR	96	0 (0.0%)	44 (45.8%)	40 (41.7%)	12 (12.5%)
4.1 mini	AR	124	31 (25.0%)	4 (3.2%)	57 (46.0%)	32 (25.8%)
5 mini	LLM	108	82 (75.9%)	0 (0.0%)	5 (4.6%)	21 (19.4%)
	ANR	92	63 (68.4%)	0 (0.0%)	2 (2.2%)	27 (29.3%)

TABLE V: Counts (percentages) of the methods used to compute the

\epsilon

threshold for the generated

\epsilon

-lexicase algorithms in Table III . Percentages are relative to the total number of

\epsilon

-lexicase algorithms generated for each LLM–configuration pairing. Only pairings that generated at least one

\epsilon

-lexicase algorithm are included.

Model	Agent	Tour.	Tournament Size
Type	Config	Total	2	3	5	7	Top-2%
oss 20b	LLM	126	8 (6.3%)	118 (93.7%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
	ANR	126	5 (4.0%)	121 (96.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
	AR	29	11 (37.9%)	18 (62.1%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
4o	LLM	126	0 (0.0%)	126 (100.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
4o	ANR	126	0 (0.0%)	125 (99.2%)	1 (0.8%)	0 (0.0%)	0 (0.0%)

TABLE VI: Counts (percentages) of tournament sizes among the algorithms classified as tournament selection in Table III . Percentages are relative to the total number of tournament selection algorithms generated for each LLM–configuration pairing. Only pairings that generated at least one tournament selection algorithm are included.

Model	Agent	Execution	Classification	Code
Type	Config	Success	Success	Similarity
oss 20b	LLM	126 (100.0%)	126 (100.0%)	0.872 [0.818, 0.904]
	ANR	126 (100.0%)	126 (100.0%)	0.890 [0.824, 0.920]
	AR	123 (97.6%)	72 (57.1%) †	0.829 [0.765, 0.881] †
4o	LLM	126 (100.0%)	126 (100.0%)	1.000 [0.971, 1.000]
	ANR	126 (100.0%)	126 (100.0%)	1.000 [0.971, 1.000]

TABLE VII: Execution success, classification success, and code similarity across LLMs and agent configurations. Success metrics are reported as counts (percentages), and code similarity as median [

Q_{1}

Q_{3}

]. A dagger (

\dagger

) indicates a significant difference from other agent configurations within the same model.

	Config	(air)	(con)	(enc)	(enh)	(hse)	(yat)
	Model &	Airfoil	Concrete	Energy Cooling	Energy Heating	Housing	Yacht
oss 20b	a LLM	604.27 [137.34, 720.23]	142.27 [130.29, 161.05]	15.08 [13.48, 17.81]	15.11 [11.94, 17.04]	52.89 [32.53, 64.01]	50.13 [20.82, 57.71]
	b ANR	189.02 [88.82, 417.27]	131.54 [108.05, 161.69]	13.24 [11.46, 15.86]	16.45 [12.46, 18.73]	51.95 [34.57, 58.63]	30.46 [21.04, 52.86]
	c AR	243.53 [106.06, 783.93]	149.67 [128.56, 199.53]	13.83 [9.89, 16.96]	16.88 [12.01, 22.57]	58.65 [35.70, 75.27]	43.71 [14.58, 126.65]
4o	a LLM	c 217.61 [110.28, 419.51]	119.57 [102.84, 141.80]	14.65 [13.09, 16.42]	c 14.96 [12.76, 16.29]	46.48 [38.90, 57.90]	c 27.34 [21.08, 31.54]
	b ANR	c 244.28 [176.07, 593.85]	145.65 [119.10, 171.28]	15.37 [13.42, 17.30]	c 15.53 [11.82, 19.24]	50.68 [42.78, 57.30]	33.93 [20.13, 45.81]

TABLE VIII: Median test-set MSE [

Q_{1}

Q_{3}

] across replicates for each problem, model type, and agent configuration; lower values indicate better performance. Values are rounded to two decimal places. Configurations are labeled a LLM, b ANR, and c AR; superscripts preceding a result indicate configurations from which that result differs significantly according to Holm-adjusted pairwise Wilcoxon rank-sum tests following a significant Kruskal–Wallis omnibus test.

Fig. 2: Test-set MSE across the six SR tasks for 5 mini–AR, tournament selection (tour), and semi-dynamic MAD

\epsilon

-lexicase selection (eps-lex). Raincloud plots show the distribution of the 21 independent replicates. Brackets denote pairwise WRS comparisons: ns (

p\geq\alpha

), * (

p<\alpha

), and *** (

p<10^{-3}

). For Energy Cooling, Housing, and Yacht, the y-axis is truncated for readability; excluded extreme values are indicated by arrows and remain included in the statistical analyses.

Genetic Programming

Genetic programming evolves programs and expression trees to discover models, heuristics, and algorithms for complex tasks.

See all 54 papers

Benchmarking Zero-Shot LLM-Generated Parent Selection in Genetic Programming for Symbolic Regression
Jul 26, 2026Hengzhe Zhang, Qi Chen, Bing Xue +2Zero-Shot Learning Large Language Model-Guided Optimization
Centre for Data Science and Artificial Intelligence & School of Engineering and Computer Science, Victoria University of Wellington, PO Box 600, Wellington 6140, New Zealand · Department of Computer Science and Engineering, Michigan State University, East Lansing, MI 48824, USA
AgenticEvolve: Evolutionary Program Search with Persistent Language Model Agents that Design and Run Experiments
Oct 22, 2026Ze-Yu Peng, Qing-Yu Lan, Zhan-He Wang +2LLMs for Evolutionary Optimization LLM-Based Program Synthesis
School of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100049, China. · International Centre for Theoretical Physics Asia-Pacific, University of Chinese Academy of Sciences, Beijing 100190, China. · Taiji Laboratory for Gravitational Wave Universe, University of Chinese Academy of Sciences, Beijing 100049, China. +2
RoboPhD: Evolving Diverse Complex Agents Under Tight Evaluation Budgets
Apr 6, 2026Andrew Borthwick, Stephen Ash, Anthony GalczakLLMs for Evolutionary Optimization Large Language Model-Guided Optimization
Independent Researchers

Jul 26, 2026J/K move · Enter open · S save

Benchmarking Zero-Shot LLM-Generated Parent Selection in Genetic Programming for Symbolic Regression

Hengzhe Zhang, Qi Chen, Bing Xue, Wolfgang Banzhaf+1

Centre for Data Science and Artificial Intelligence & School of Engineering and Computer Science, Victoria University of Wellington, PO Box 600, Wellington 6140, New Zealand · Department of Computer Science and Engineering, Michigan State University, East Lansing, MI 48824, USA

Zero-Shot Learning Large Language Model-Guided Optimization Genetic Programming LLM-Based Program Synthesis Symbolic Regression

Open paper AskPDF arXiv

Parent selection significantly affects exploration, exploitation, and complexity control in genetic programming (GP) for symbolic regression. It is unclear whether large language models (LLMs) can synthesize effective operators in a zero-shot setting without iterative meta-evolution. Here, zero-shot means that the model receives only the task description, with no reference operators or iterative feedback. In this work, we benchmark zero-shot synthesis of parent-selection operators across eight LLMs within a standard GP framework for symbolic regression. Each model receives the same natural-language prompt to generate a parent-selection operator, which is then evaluated in a standard GP framework with only the parent-selection operator replaced, while all other components and the evolutionary-search budget are held constant. For each LLM, ten independent zero-shot operators are evaluated on twelve OpenML regression benchmarks and compared against automatic lexicase and tournament selection baselines. Claude Sonnet~~4.6 and Gemini~~3.1 Pro stand out for consistently strong performance on both training and held-out test

R^2

. The strongest operator in our benchmark---a Kimi~K2.5 zero-shot synthesis---surpasses the automatic lexicase and tournament baselines in search effectiveness. These results suggest that zero-shot LLM synthesis is a viable approach to generating competitive GP selection operators. Analysis shows that many generated operators use semantics to guide selection, suggesting that LLMs can produce non-trivial search heuristics from the task description alone. We also examine the relationship between public LLM leaderboard rankings and GP performance. Widely used benchmarks, such as Humanity's Last Exam and SWE-bench Verified, strongly correlate with training

R^2

, while their relationship to held-out test

R^2

is weaker and less clear.

Abstract

Genetic Programming

Explore similar work

Benchmarking Zero-Shot LLM-Generated Parent Selection in Genetic Programming for Symbolic Regression

AgenticEvolve: Evolutionary Program Search with Persistent Language Model Agents that Design and Run Experiments

RoboPhD: Evolving Diverse Complex Agents Under Tight Evaluation Budgets