🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

"Impact of Training Dataset Size on Neural Answer Selection Models" w/ Trond Linjordet accepted as short paper at #ecir2019

ICTIR'20 program is ready, we have 2 keynotes, 6 industry talks and more see ictir2020.org/program-overvi…. We will be using a cool virtual platform (gather.town) for the conference to join the talks and discussions register now! ictir2020.org/participation/

“Would You Ask it that Way? Measuring and Improving Question Naturalness for Knowledge Graph Question Answering” w/ Trond Linjordet accepted as a resource paper at #sigir2022

Trond Linjordet Presenting our work on novel evaluation measures for KGQA. at ICTIR 2024 SIGIR 2025 #SIGIR2022 IAI group Universitetet i Stavanger

New preprint! We developed a SMILES-based encoding of M–L bonds to optimize a generative VAE model yielding metal ligands with high validity and novelty, which can also be conditioned by the properties of the associated homoleptic complexes in a XAI way 👉 go.shr.lc/4dX8XnX

Interested on deep generative AI for metal complexes? Check out the updated preprint on our JT-VAE model, now conditioning on both whole-complex and free-ligand properties 👉 doi.org/10.26434/chemr…

Now out in JACS Au, our inverse design model for transition metal chemistry, supporting generative tasks conditioned by ligand and metal complex properties. We implemented dual-objective generation and it was fascinating to see navigation in latent space! pubs.acs.org/doi/10.1021/ja…