HepScript: A Dual-Use DSL for Human-AI Collaborative Data Analysis Workflows in High-Energy Physics

Junkun Jiao , Tong Liu , Ke Li , WeiMin Song , Yipu Liao , Bolun Zhang , Beijiang Liu , Chang-Zheng Yuan

show 1 more author

Yue Sun

Authors on Pith no claims yet

Pith reviewed 2026-05-09 13:42 UTC · model grok-4.3

classification ✦ hep-ex cs.AIcs.MA

keywords analysishepscriptautomationdataworkflowsagentscodecollaborative

0 comments

The pith

HepScript is a dual-use DSL that abstracts HEP analysis logic into a constrained syntax, reducing human-written code by 93% and enabling AI agents to generate executable analysis specifications from literature with 95% success.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

High-energy physics experiments generate vast datasets that demand intricate analysis using experiment-specific software stacks. The authors created HepScript, a specialized programming language with a limited set of rules that lets physicists express analysis steps in a clear, high-level way. This same language is structured so AI systems can reliably produce correct code from it. In tests on the BESIII experiment, the approach cut the amount of code that humans needed to write by 93 percent. AI agents could read published papers and turn the described analysis steps into working, executable specifications 95 percent of the time. The limited grammar creates a manageable space of possible actions, which helps AI avoid common errors that occur when trying to generate arbitrary code.

Core claim

HepScript's constrained grammar defines a tractable action space, enabling AI agents to autonomously generate executable specifications for core analysis stages directly from published literature with a 95% success rate. In our case studies, this abstraction reduces the required human-written code by 93%.

Load-bearing premise

That the constrained syntax of HepScript remains sufficiently expressive to capture the full complexity of real HEP analysis workflows without requiring frequent extensions or losing critical domain-specific details.

read the original abstract

The escalating data scale in High-Energy Physics (HEP) fuels a growing aspiration for higher analytical efficiency. While Large Language Models (LLMs) offer a path toward automation via agentic AI, they struggle with complex scientific workflows that require deep domain knowledge and are tightly coupled to experiment-specific codebases. To address this, we introduce a methodology centered on HepScript, a dual-use Domain-Specific Language (DSL) for HEP data analysis workflows. HepScript serves as a shared formal interface, abstracting HEP analysis logic into a constrained syntax that is both intuitive for human experts and reliably generable by AI agents. First developed for the Beijing Spectrometer III (BESIII) experiment, HepScript hides the complexity of the underlying software stack, translating high-level analysis intent into low-level, production-ready code. In our case studies, this abstraction reduces the required human-written code by 93\%. Crucially, HepScript's constrained grammar defines a tractable action space, enabling AI agents to autonomously generate executable specifications for core analysis stages directly from published literature with a 95\% success rate. Our work demonstrates a scalable pathway toward human-AI collaborative systems, where a formally specified DSL acts as an unambiguous translation layer between human expertise, AI automation, and production environment, rendering previously intractable automation problems solvable.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces HepScript, a dual-use Domain-Specific Language (DSL) for High-Energy Physics (HEP) data analysis workflows, initially developed for the BESIII experiment. It positions HepScript as a constrained syntax that abstracts complex analysis logic, enabling both human experts and AI agents to interact with production codebases. The central claims are a 93% reduction in human-written code and a 95% success rate for AI agents generating executable specifications directly from published literature in case studies.

Significance. Should the reported quantitative results be substantiated with rigorous methodology, this approach could offer a valuable framework for integrating AI into HEP analysis pipelines, potentially improving efficiency in handling large datasets. The idea of a DSL as an unambiguous interface between humans, AI, and experiment-specific software is promising for collaborative systems. However, without details on how the success rates were measured or the representativeness of the case studies, the significance remains provisional.

major comments (2)

[Abstract] Abstract: The claims of a 93% reduction in human-written code and 95% AI success rate in autonomously generating executable specifications from published literature are presented without any definition of success criteria (e.g., syntactic validity, semantic fidelity to paper intent, or end-to-end executability), number of trials, inter-annotator agreement, baseline comparisons, or breakdown by analysis stage such as event selection or fitting. This directly prevents evaluation of the central claim that the constrained grammar defines a tractable yet sufficiently expressive action space.
[Abstract] Abstract: No evidence is provided that the HepScript grammar remains expressive enough to capture the full complexity of real BESIII-style workflows (including systematic evaluations or experiment-specific details) without frequent extensions, which is load-bearing for the assertion that the DSL renders previously intractable automation problems solvable.

minor comments (1)

The abstract refers to 'case studies' and 'our work' without cross-references to specific sections, tables, or figures that would detail the evaluated literature excerpts or code reduction measurements.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on the domain assumption that HEP workflows can be usefully abstracted into a constrained syntax and that this syntax is sufficient for both human use and reliable AI generation.

axioms (1)

domain assumption HEP analysis workflows can be abstracted into a constrained syntax without loss of essential functionality.
This premise is required for HepScript to serve as a complete translation layer between high-level intent and production code.

invented entities (1)

HepScript DSL no independent evidence
purpose: To act as a shared formal interface that abstracts HEP analysis logic for both humans and AI agents.
HepScript is the core new artifact introduced by the paper.

pith-pipeline@v0.9.0 · 5566 in / 1311 out tokens · 76453 ms · 2026-05-09T13:42:27.502234+00:00 · methodology

discussion (0)

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Toward a Community Roadmap for High Energy Physics and Artificial Intelligence in China and Beyond
hep-ph 2026-05 unverdicted novelty 2.0

A workshop-derived partial overview of AI applications in high energy physics serves as a starting point for building a community roadmap.