This work builds four modules (for drug-discovery chemistry, materials science, machine learning, and mineral geochemistry) in the ArticleMiner framework, and evaluates them on 163 papers, including a new geochemistry benchmark with expert-curated ground truth.
Md Abrar Jahin, Craig A. Knoblock, Jay Pujara· arXiv (Cornell University)· 1 citation
QUEST, which adds no trainable parameters to the standard confidence-distribution learning pipeline, improves confidence prediction and link prediction on six of eight metric-dataset pairs over prior methods and matches the previous best on the remaining two, while removing the instability spike observed on dense graph...
Md Abrar Jahin, Taufikur Rahman Fuad, Jay Pujara et al.· 0 citations
The results characterize the expressive power of global-attention GFMs and provide an encoder-agnostic diagnostic for detecting 1-WL-induced representation equivalence.
Md Abrar Jahin, Craig A. Knoblock, Jay Pujara· arXiv.org· 0 citations
KB-Anchored Rule Induction (KARI), a pipeline based on LLMs that generates affordance rules grounded in evidence from knowledge bases (KBs), is developed, which improves open-source LLMs by up to 15.2 points, while the limited coverage of KBs hinders further gains.
Yifan Jiang, Mei-Ge Yang, Zitong Li et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
