Search papers, labs, and topics across Lattice.
Research Institute of Trustworthy Autonomous Systems, Department of Computer Science and Engineering, Southern University of Science and Technology, Southern University of Science and Technology Shenzhen
3
0
7
WebCQ achieves 33.3% more state exploration and 42.2% more unique actions than previous MARL methods, revolutionizing web GUI testing scalability.
English-only post-training of LLMs is suboptimal: even a single non-English language improves both English performance and cross-lingual generalization.
JSON tool-calling is so fragile it's practically unusable: BloClaw offers a 98.8% success rate with its novel XML-Regex Dual-Track Routing Protocol.
