TiEBe: Tracking Language Model Recall of Notable Worldwide Events Through Time
Thales Sales Almeida, Giovana Kerche Bonás, João Guilherme Alves Santos, Hugo Abonizio, Rodrigo Nogueira · 2025
Read TiEBe: Tracking Language Model Recall of Notable Worldwide Events Through Time by Thales Sales Almeida, Giovana Kerche Bonás, João Guilherme Alves Santos, Hugo Abonizio, Rodrigo Nogueira (2025) 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-2501-07482.txt.
Description
Thales Almeida and four co-authors build a benchmark of world events in this two thousand twenty-five paper. Large language models are now everywhere, yet their grip on recent events is spotty. Existing tests check general facts, not how recall fades with time or varies by place. Their benchmark, TiEBe, holds more than twenty-three thousand question-answer pairs. The events span over ten years and twenty-three regions of the world.
Each item centers on a notable event, large or local. Models can then be scored on how well they recall events by era and region. The authors use it to track where model memory is strong and where it thins. Such a tool helps teams decide when a model needs fresh training. It also warns users about stale answers on current affairs.
Catalog record
| CREATOR | Thales Sales Almeida, Giovana Kerche Bonás, João Guilherme Alves Santos, Hugo Abonizio, Rodrigo Nogueira |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-07482