Yu Su

This work proposes DualSQL, a new Text-to-SQL system consisting of two agents powered by a single model backbone, enabling joint optimization through a robust multi-agent reinforcement learning (RL) framework, and introduces a new SQL correctness metric, robust execution match (REX), to more accurately judge SQL correctness and assign reward signals.

GEAR slashes inference time by up to 2866 times while boosting AUC scores beyond conventional supervised models.

Achieving a 90.3% success rate in corridor navigation, this framework outperforms traditional methods by leveraging shared voxel-map representations for multi-UAV cooperation.

Local optima can be proactively escaped, leading to a 30% increase in collaboration success rates among UAVs in complex navigation tasks.

AgenticDataBench reveals that LLM-based data agents can be rigorously evaluated across diverse real-world scenarios, highlighting their strengths and weaknesses in handling complex data tasks.

Current AI agents only manage to complete 20.6% of complex real-world tasks, revealing a stark gap in their capabilities compared to human users.

A unified assessment framework reveals hidden insights about agent performance, transforming how we evaluate AI systems.

Current agents are alarmingly susceptible to skill-based attacks, with success rates reaching over 86%, exposing a critical vulnerability in AI safety.