What question did this study set out to answer?

This research aims to distinguish between different types of variations in code proposals generated by a stochastic code generator and their impact on execution outcomes.

June 1, 2026Open Access

Proposal Collapse and Execution Fibers in Stochastic Program Generation

View Full Paper

Ask AI

Bookmark

Authors

TDThomas Dionysopoulos

Discussion

Member takes

Overview

Randomized trial investigates execution variability in code proposals from natural-language prompts, suggesting deterministic semantics shapes proposal structures.

Key Points

This research aims to distinguish between different types of variations in code proposals generated by a stochastic code generator and their impact on execution outcomes.
Conducted controlled perturbation experiments on 2,200 code proposals (1,200 baseline, 1,000 perturbed) to analyze execution classes.
Formulated a measurement system using execution fibers to classify execution variations.
Examined the correlation between prompt support and fiber cardinality.
Identified 28 execution classes with non-uniform fiber cardinality (Gini = 0.467).
Observed near-perfect stability in proposals with synonym perturbations (sigma = 0.992, rho = 0.985).
Established a strong correlation (r = 0.925) between prompt support and fiber cardinality.

Cite This Study

Thomas Dionysopoulos (2026) studied this question.

synapsesocial.com/papers/6a1d22f702fbce9130638b21 https://doi.org/10.5281/zenodo.20457989

View Full Paper

Ask AI

Bookmark

June 1, 2026Open Access

Proposal Collapse and Execution Fibers in Stochastic Program Generation

View Full Paper

Ask AI

Bookmark

Authors

TDThomas Dionysopoulos

Discussion

Member takes

Overview

Randomized trial investigates execution variability in code proposals from natural-language prompts, suggesting deterministic semantics shapes proposal structures.

Key Points

This research aims to distinguish between different types of variations in code proposals generated by a stochastic code generator and their impact on execution outcomes.
Conducted controlled perturbation experiments on 2,200 code proposals (1,200 baseline, 1,000 perturbed) to analyze execution classes.
Formulated a measurement system using execution fibers to classify execution variations.
Examined the correlation between prompt support and fiber cardinality.
Identified 28 execution classes with non-uniform fiber cardinality (Gini = 0.467).
Observed near-perfect stability in proposals with synonym perturbations (sigma = 0.992, rho = 0.985).
Established a strong correlation (r = 0.925) between prompt support and fiber cardinality.

Cite This Study

Thomas Dionysopoulos (2026) studied this question.

synapsesocial.com/papers/6a1d22f702fbce9130638b21 https://doi.org/10.5281/zenodo.20457989

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark