Tag
The paper introduces aDSL, a co-designed domain-specific language and multi-agent system that improve LLM-driven 3D program synthesis through relational operators and iterative feedback, enhancing robustness and controllability in 3D content creation.
AutoPrune is a training-free framework that uses LLMs to automatically design visual-token pruning policies for multimodal LLMs via a domain-specific language and residual search, achieving high efficiency with minimal performance loss (99% performance retained while removing 94.4% of visual tokens).
This paper introduces ADSL-PDE, a domain-specific language that provides a structured search space for auto-designing neural PDE solvers, improving search efficiency and optimization stability by abstracting away low-level implementation details. The evolutionary agent built on this representation achieves over 52% performance improvement within the first ten iterations across PDE benchmarks.
PULSE is a new executable contract language for spatiotemporal knowledge graph engineering, providing a typed runtime with role-based write effects, safety properties verified in Lean 4, and trace parity across large datasets.
MorphoHDL is a minimalistic domain-specific language for designing and growing circuits, offering a simple syntax for hardware description.
Forethought is a neurosymbolic reasoning system that treats reasoning as an explicit, verifiable program composed from symbolic and neural primitives. It improves base-model accuracy by about 30% relative and enables small models to match frontier models while being model-agnostic and auditable.
VeryTrace is a zero-shot verification-and-repair framework that formalizes LLM reasoning traces into a compilable representation using a DSL, enabling step-level error localization through a hybrid of deterministic checks and LLM audits. It improves accuracy across math, robotics, and relational reasoning without domain-specific training.
PromptMN is a domain-specific language that adds typed directives to natural language prompts to reduce ambiguity. It is evaluated on frontier models and shows feasibility for clearer human-AI interaction.
This article explores the design of Lispy domain-specific languages using SCSS, a Scheme-based CSS preprocessor, as a case study. It discusses how SCSS represents CSS as first-class values and the limitations of its implementation.
The author details the process of designing a custom query language tailored for non-technical analysts to filter vehicle maintenance data, outlining user needs, data schema, and specific use cases.