Search papers, labs, and topics across Lattice.
Guangdong University of Technology
5
0
7
CDFM outperforms traditional causal discovery algorithms by leveraging a unified framework that adapts to diverse datasets without the need for extensive retraining.
None of the 30 LLM agents evaluated in CausalGame demonstrated reliable causal thinking, revealing a critical gap in AI's ability to perform scientific reasoning.
MoVA achieves superior video-text alignment by disentangling evolving visual concepts from static textual descriptions, outperforming existing models in handling long sequences.
User feedback prediction can be learned, revealing critical performance gaps in AI assistants that traditional evaluation methods overlook.
Forget static rubrics and expensive external models: EvoRubric co-evolves a single policy to generate both responses and the rubrics to evaluate them, outperforming traditional RLHF methods in open-ended generation tasks.
