Dynamic Benchmarking for Information Seeking Agents
iAgentBench
A dynamic benchmark for cross-source sensemaking in information-seeking agents on high-traffic real-world topics.
Overview of the iAgentBench Pipeline
Important links
Access the code, dataset, and resources.
Citation
If you use iAgentBench, please cite our paper.
@article{dammu2026iagentbench,
title = {iAgentBench: Benchmarking Sensemaking Capabilities of Information-Seeking Agents on High-Traffic Topics},
author = {Dammu, Preetam Prabhu Srikar and Palkhiwala, Arnav and Roosta, Tanya and Shah, Chirag},
journal = {arXiv preprint},
year = {2026},
eprint = {2603.04656},
archivePrefix = {arXiv},
primaryClass = {cs.CL},
url = {https://arxiv.org/abs/2603.04656}
}
Contact
Get in touch.