Search papers, labs, and topics across Lattice.
4
6
5
FLARE-AI transforms the fragmented AI flaw reporting landscape by enabling a single report to reach multiple stakeholders, enhancing collaboration and speeding up remediation efforts.
Predictions alone can mislead; understanding how they reshape decision-making is crucial for effective automated systems.
A global consensus on AI safety risks and capabilities has emerged from a panel of 100+ independent experts, representing a landmark effort in international collaboration.
Despite progress in AI agent capabilities, reliability across crucial dimensions like consistency and robustness remains stubbornly low, revealing a critical gap in current evaluation practices.
