Sliceformer reformulates static program slicing as seq2seq using CodeT5+ with dataflow-aware pretraining via DFG permutation and span corruption plus constrained decoding, yielding up to 22% ExactMatch gains on Java and Python benchmarks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Static Program Slicing Using Language Models With Dataflow-Aware Pretraining and Constrained Decoding
Sliceformer reformulates static program slicing as seq2seq using CodeT5+ with dataflow-aware pretraining via DFG permutation and span corruption plus constrained decoding, yielding up to 22% ExactMatch gains on Java and Python benchmarks.