JSON

The 'JSON' dataset/benchmark contains structured data formats used to evaluate Large Language Models' (LLMs) capabilities in generating and converting structured outputs while assessing their structural fidelity and correctness.

Papers using JSON (7)

StructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs2025 · 1 citesStructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs2025 · 1 citesInferring Input Grammars from Code with Symbolic Parsing2025 · 1 citesCottontail: Large Language Model-Driven Concolic Execution for Highly Structured Test Input Generation2025SynCode: LLM Generation with Grammar Augmentation2024 · 3 citesV-Star: Learning Visibly Pushdown Grammars from Program Inputs2024 · 3 citesGuiding Large Language Models to Generate Computer-Parsable Content2024 · 1 cites