Search papers, labs, and topics across Lattice.
SKLP, Institute of Computing Technology, Chinese Academy of Sciences
2
0
4
0
Ghost references in LLM-generated code can be eliminated by constraining generation to valid runtime environments, ensuring both grammatical and semantic correctness.
LLM code generation benchmarks are likely overestimating model capabilities: adversarial test suite scaling reveals substantial weaknesses in even state-of-the-art models.
