Publications
Conference Papers
Alireza Dizaji, Benedict Aaron Tjandra, Mehrab Hamidi, Shenyang Huang, and Guillaume Rabusseau
KDD 2026 · Data & Benchmarks (Oral)
Also presented at the KDD 2025 MLoG-GenAI Workshop (Oral).
A synthetic benchmark that isolates temporal reasoning skills and exposes limitations in temporal graph neural networks.
Workshop Papers
Lucio M. Dery*, Benedict Aaron Tjandra*, Siavash Samiei, Adhiguna Kuncoro, Zohar Yahav, Jiajun Shen, and Arthur Szlam
* Equal contribution.
COLM 2026 LLA Workshop
We combine learned skills and textual knowledge to synthesize new skills through instruction-steered prefix generation.
Benedict Aaron Tjandra, Avi Wigderson, João G. M. Araújo, Alex Vitvitskyi, Federico Barbero, and Petar Veličković
ICML 2026 FoGen and GFM Workshops
We establish backward mixing time as a bound on information fidelity, revealing a speed–fidelity trade-off in symmetric feedforward computational graphs.
Benedict Aaron Tjandra, Federico Barbero, and Michael Bronstein
NeurIPS Symmetry and Geometry in Neural Representations Workshop 2024 (Oral)
We show that adding source-target identification to Temporal Graph Networks (TGNs) strictly increases its expressivity and demonstrate that the resulting architecture outperforms all TG methods in dynamic node affinity prediction on all TGB datasets.
Benedict Aaron Tjandra, Muhammed Razzak, Jannik Kossen, and Yarin Gal
NeurIPS Safe Generative AI Workshop 2024
We fine-tune LLMs using semantic entropy to make them abstain from answering uncertain questions.
Preprints
Edward Elson Kosasih, Joaquin Cabezas, Xavier Sumba, Piotr Bielak, Kamil Tagowski, Kelvin Idanwekhai, Benedict Aaron Tjandra, and Arian Rokkum Jamasb
arXiv · 2021
We conducted experiments using graph neural network ensembles to predict molecular HOMO-LUMO gaps.