01 / MOTIVATION
Why this work matters
Numerical claims can be deceptively difficult to verify, since small differences in quantities, units and context can change the verdict.
A fine-tuned LLM approach for numerical claim verification that placed second in the shared task.
02 / METHODOLOGY
Research approach
The shared-task system uses fine-tuned language models to verify numerical statements; it ranked second in its CheckThat! task.
03 / CONTRIBUTIONS
Main contribution
Fine-tuned LLMs for verification of numerical claims.
- 2nd place in CheckThat! Lab 2026 Task 2
Consult the complete publication for numerical findings and experimental qualifications.
04 / PUBLICATION
Read the paper
Diana Zhang, Xiaoying Song, Mohotarema Rashid, Anirban Saha Anik, Junhua Ding, and Lingzi Hong. (2026). VeriFind at CheckThat! 2026: Enhancing Numerical Claim Verification via Fine-Tuned LLMs. Working Notes of CLEF 2026, CEUR Workshop Proceedings, Vol. 4283, pp. 1361–1369.