Referenceless metrics (e.g., CLIPScore) use pretrained vision–language
m...
Inductive reasoning is a core problem-solving capacity: humans can ident...
In recent years, deep learning-based solar forecasting using all-sky ima...
Language model training in distributed settings is limited by the
commun...
Language models often achieve higher accuracy when reasoning step-by-ste...
Token embeddings, a mapping from discrete lexical symbols to continuous
...
Despite recent success in large language model (LLM) reasoning, LLMs sti...
Language models (LMs) are becoming the foundation for almost all major
l...
Few images on the Web receive alt-text descriptions that would make them...
Generating step-by-step "chain-of-thought" rationales improves language ...
Advancing probabilistic solar forecasting methods is essential to suppor...
Learning disentangled representations is regarded as a fundamental task ...
While machine learning models capable of producing uncertainty estimates...
A significant challenge in developing AI that can generalize well is
des...
This paper introduces a simple and explicit measure of word importance i...