Search papers, labs, and topics across Lattice.
James
8
0
8
17
Brevis compresses tensor data by synthesizing a DSL program, achieving over 30% smaller archives than leading general-purpose compressors while ensuring bit-exact reconstruction.
Current LLMs achieve only a 12.30% success rate in generating executable scientific code, highlighting a significant gap in their capabilities.
As SE agents become mainstream, the shift to evaluation-driven development reveals new challenges that could redefine engineering workflows.
LLMs can now automatically fix critical security vulnerabilities in Trusted Execution Environments with high success rates, even without standardized development guidelines.
Skip the costly TEE setup: SymTEE uses LLMs to automatically create mock environments for symbolic execution, slashing the complexity of finding vulnerabilities in trusted computing systems.
Turns out, teaching LLMs to *think* like reverse engineers beats just throwing more parameters at the problem of binary deobfuscation.
Run code LLMs 10x faster and with 6x less memory on your laptop without sacrificing accuracy, thanks to a new quantization and compilation approach.
Automating ESG reporting with LLM-powered agents transforms it from a static compliance exercise into a dynamic, data-driven system for sustainability governance.
