Yuxiong He

AI-native SQL queries expose critical gaps in model performance, with top proprietary models still struggling to achieve 70% execution accuracy.

Training LLMs without ground-truth solutions can yield significant performance improvements, as shown by RiVER's success in enhancing both score-based and exact-solution benchmarks.

SNAS enables secure, efficient external communication in multi-tenant environments without compromising isolation or performance.

Efficiently navigating hierarchical memory can cut token usage by over 77% while boosting task performance in LLMs.

Text-to-SQL models can now achieve significantly higher accuracy by grouping and ranking SQL candidates based on execution results, then strategically resampling when the initial pool is lacking.

Data skew can cripple Snowpark UDF performance, but DySkew's dynamic redistribution slashes execution time and boosts resource utilization in real-world workloads.

Stop hand-crafting hints for RL agents: HiLL learns to generate adaptive hints that actually improve the agent's performance on the original task, not just the hinted one.

Forget prompt engineering – LSE trains LLMs to self-edit their own contexts at test time, outperforming even GPT-5 and Claude Sonnet 4.5 in Text-to-SQL and question answering.