Search papers, labs, and topics across Lattice.
1
0
3
0
Forget massive models: surprisingly small, specialized LLMs can achieve state-of-the-art decompilation of x86-64 assembly into modern languages like Dart, rivaling the performance of models 100x larger.