Search papers, labs, and topics across Lattice.
1
0
3
2
Existing QA benchmarks are too easy for LLMs, so iAgentBench offers a more realistic challenge by requiring agents to synthesize information from multiple sources on high-traffic topics.
