Am I Blue or Is My Hobby Counting the Teardrops? Expression Leakage in Large Language Models as a Symptom of Irrelevancy Disruption

Berkay Kopru, Mehrzad Mashal, Yigit Gurses, Akos Kadar, Maximilian Schmitt, Ditty Mathew, Felix Burkhardt, Florian Eyben, Björn W. Schuller


Abstract
Large language models (LLMs) have advanced natural language processing (NLP) skills such as through next-token prediction and self-attention, but their ability to integrate broad context also makes them prone to incorporating irrelevant information. Prior work has focused on semantic leakage—bias introduced by semantically irrelevant context.In this paper, we introduce expression leakage, a novel phenomenon where LLMs systematically generate sentimentally charged expressions that are semantically unrelated to the input context. To analyse the expression leakage, we collect a benchmark dataset along with a scheme to automatically generate a dataset from free-form text from common-crawl. In addition, we propose an automatic evaluation pipeline that correlates well with human judgment, which accelerates the benchmarking by decoupling from the need of annotation for each analysed model. Our experiments show that, as the model scales in the parameter space, the expression leakage reduces within the same LLM family. On the other hand, we demonstrate that expression leakage mitigation requires specific care during the model building process, and cannot be mitigated by prompting. In addition, our experiments indicate that, when negative sentiment is injected in the prompt, it disrupts the generation process more than the positive sentiment, causing a higher expression leakage rate.
Anthology ID:
2025.ranlp-1.68
Volume:
Proceedings of the 15th International Conference on Recent Advances in Natural Language Processing - Natural Language Processing in the Generative AI Era
Month:
September
Year:
2025
Address:
Varna, Bulgaria
Editors:
Galia Angelova, Maria Kunilovskaya, Marie Escribe, Ruslan Mitkov
Venue:
RANLP
SIG:
Publisher:
INCOMA Ltd., Shoumen, Bulgaria
Note:
Pages:
578–586
Language:
URL:
https://aclanthology.org/2025.ranlp-1.68/
DOI:
Bibkey:
Cite (ACL):
Berkay Kopru, Mehrzad Mashal, Yigit Gurses, Akos Kadar, Maximilian Schmitt, Ditty Mathew, Felix Burkhardt, Florian Eyben, and Björn W. Schuller. 2025. Am I Blue or Is My Hobby Counting the Teardrops? Expression Leakage in Large Language Models as a Symptom of Irrelevancy Disruption. In Proceedings of the 15th International Conference on Recent Advances in Natural Language Processing - Natural Language Processing in the Generative AI Era, pages 578–586, Varna, Bulgaria. INCOMA Ltd., Shoumen, Bulgaria.
Cite (Informal):
Am I Blue or Is My Hobby Counting the Teardrops? Expression Leakage in Large Language Models as a Symptom of Irrelevancy Disruption (Kopru et al., RANLP 2025)
Copy Citation:
PDF:
https://aclanthology.org/2025.ranlp-1.68.pdf