A custom LLM agent achieves 94% manually verified success on a new benchmark of 35 software analysis setups, outperforming baselines at 77%, but struggles with stage mixing, error localization, and overestimating its own success.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.SE 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Evaluating LLM Agents on Automated Software Analysis Tasks
A custom LLM agent achieves 94% manually verified success on a new benchmark of 35 software analysis setups, outperforming baselines at 77%, but struggles with stage mixing, error localization, and overestimating its own success.