Event Detection with a Context-Aware Encoder and LoRA for Improved Performance on Long-Tailed Classes

Abdullah Al Monsur, Nitesh Vamshi Bommisetty, Gene Louis Kim · 2026

Read Event Detection with a Context-Aware Encoder and LoRA for Improved Performance on Long-Tailed Classes by Abdullah Al Monsur, Nitesh Vamshi Bommisetty, Gene Louis Kim (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-2601-11932.txt.

Description

Abdullah Al Monsur, Nitesh Vamshi Bommisetty, and Gene Louis Kim improve event spotting in text for rare event kinds. Two flaws drive their work. First, models that read text in one way miss the rich two-way context that event spotting needs.

They add a context-aware encoder to fix that flaw. They also use LoRA, a light way to tune a model, to aid rare classes. Rare events have few training cases, so they are easy to miss. Better rare class skill helps news and risk tools that track events. Rare class gains matter most where events are few and grave.

Catalog record

CREATORAbdullah Al Monsur, Nitesh Vamshi Bommisetty, Gene Louis Kim
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-11932

Browse the full library · About the Tower of Babel