Search papers, labs, and topics across Lattice.
University of Waterloo
5
0
9
3
Language models struggle to predict changes in dynamic environments, achieving only 13.8% accuracy in a rigorous benchmark designed to test causal reasoning and state transitions.
Traditional recommendation algorithms falter when faced with LLM agents, revealing a surprising shift from personalization to mere structural pattern matching.
MVOFormer outperforms traditional learning-based MVO methods by achieving superior zero-shot generalization and robustness through a novel flow-semantic architecture.
SPECTRA reveals that Scratch programs can be behaviorally equivalent despite significant syntactic differences, challenging traditional notions of program comparison.
Grading Scratch programs just got easier: Raven uses LLMs to watch student-created videos and automatically assess whether they meet the assignment's goals.
