Search papers, labs, and topics across Lattice.
7
0
10
1
Safety in education-facing LLMs is inversely related to their teaching effectiveness, challenging assumptions about model specialization.
Interactive dialogue can unlock creative potential that static assessments overlook, leading to richer evaluations of creativity in AI contexts.
LLMs can significantly enhance their collaborative performance, achieving over 24% better engagement with human partners when trained in realistic game scenarios.
Multi-agent prompt refinement can significantly boost text-to-video generation quality in complex scenarios, exceeding SOTA baselines by up to 3.28% on standard benchmarks.
LLMs can now generate coherent, diagram-rich explanations for K-12 STEM problems with high accuracy, opening new avenues for automated educational content creation.
Fine-tuning language models on role-specific representations bridges the semantic gap in cognitive diagnosis, substantially boosting performance across diverse educational tasks.
Forget hand-crafted student profiles: HACHIMI generates a million theory-aligned, quota-controlled student personas, offering a standardized synthetic population for educational LLM benchmarking and social-science simulations.
