Jiahao Yuan

Hallucinations in video models can be mitigated by intelligently patching visual tokens, leading to a 50.60% improvement in response accuracy.

LLM multi-agent systems can achieve significantly higher accuracy at a fraction of the cost by learning to selectively delegate tasks instead of relying on rigid orchestration.

LLMs can be adapted for scenario-specific user representation by conditioning on queries, achieving state-of-the-art performance on Alipay benchmarks and demonstrating practical effectiveness in online A/B testing.