A Benchmark for Deep Information Synthesis
Debjit Paul, Daniel Murphy, Milan Gritta, Ronald Cardenas, Victor Prokhorov, Jun Wang, Gerasimos Lampouras Dataset Contributors, Lena Sophia Bolliger, Aysim Toker, Roy Miles, Andreea-Maria Oncescu, Jasivan, Alex Sivakumar, Philipp Borchert, Ismail Elezi, Meiru Zhang, Ka Yiu Lee, Guchun Zhang · 2026
Read A Benchmark for Deep Information Synthesis by Debjit Paul, Daniel Murphy, Milan Gritta, Ronald Cardenas, Victor Prokhorov, Jun Wang, Gerasimos Lampouras Dataset Contributors, Lena Sophia Bolliger, Aysim Toker, Roy Miles, Andreea-Maria Oncescu, Jasivan, Alex Sivakumar, Philipp Borchert, Ismail Elezi, Meiru Zhang, Ka Yiu Lee, Guchun Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-21143.txt.
Description
Language-model agents increasingly tackle research-like tasks, but benchmarks rarely test synthesis across sources. Debjit Paul and sixteen co-authors introduce such a benchmark in this 2026 arXiv paper: DEEPSYNTH, with one hundred twenty tasks across seven domains covering sixty-seven countries. Annotators gathered official sources, formed hypotheses, analyzed manually, and designed tasks with verifiable answers.
Eleven leading models and deep-research agents score at most just under nine on F1 and seventeen and a half on a judge metric, showing how hard synthesis remains. Agents struggle with hallucinations and reasoning over large information spaces. The benchmark gives the field a realistic yardstick for guiding research beyond simple fact retrieval.
Catalog record
| CREATOR | Debjit Paul, Daniel Murphy, Milan Gritta, Ronald Cardenas, Victor Prokhorov, Jun Wang, Gerasimos Lampouras Dataset Contributors, Lena Sophia Bolliger, Aysim Toker, Roy Miles, Andreea-Maria Oncescu, Jasivan, Alex Sivakumar, Philipp Borchert, Ismail Elezi, Meiru Zhang, Ka Yiu Lee, Guchun Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-21143