BLAST is a new benchmark using ten graph problems and two semantic metrics to evaluate LLMs on generating correct ASP code.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
BLAST: Benchmarking LLMs with ASP-based Structured Testing
BLAST is a new benchmark using ten graph problems and two semantic metrics to evaluate LLMs on generating correct ASP code.