Daming Li

Traditional recommendation algorithms falter when faced with LLM agents, revealing a surprising shift from personalization to mere structural pattern matching.

Grading Scratch programs just got easier: Raven uses LLMs to watch student-created videos and automatically assess whether they meet the assignment's goals.

Multimodal repair isn't always worth the cost: selectively escalating to multimodal prompting based on runtime signals in Scratch yields a better success-cost-energy tradeoff than uniformly applying it.