Search papers, labs, and topics across Lattice.
7
0
12
0
R2S-Eval reveals that automated evaluation can outperform manual success counting by capturing nuanced differences in robot behavior quality.
Evolving data snapshots can boost LLM performance by over 20% in both intrinsic and extrinsic evaluations, reshaping how we approach human-centered AI alignment.
Newer LLMs may ace build tasks, but they struggle with delivering correct app behaviors, especially when specifications are involved.
Safety in education-facing LLMs is inversely related to their teaching effectiveness, challenging assumptions about model specialization.
Zero-loss exactness in optimal transport dynamics could revolutionize how we approach matching problems in high-dimensional spaces.
Integrating a decision-making stage into radiology report generation can dramatically enhance diagnostic accuracy by resolving conflicting information without retraining models.
Focusing optimization efforts on the Decomposer can yield a 19% improvement in proof autoformalization efficiency without increasing costs.
