Evaluation Awareness Shifts from Format to Context with Model Scale

Abstract:Evaluation awareness poses an unprecedented threat to model evaluation, but the mechanisms by which models detect it remain unknown. This study focuses on determining this and identifying contrasting mechanisms between smaller and larger models. While smaller models use the prompt's format sensitivity to detect evaluation, larger models often rely on higher-order reasoning to detect it. We evaluated Gemma 3 (1B, 4B, and 12B), Phi-3 (Mini and Medium), and Llama-3 8B using Chain-of-Thought analysis, representation probing, and Integrated Gradients attribution. Motivated by these findings, we propose a dual-pathway intervention that combines prompt sanitization with activation counter-steering to suppress both external evaluation triggers and their internal representations. Across 200 highly evaluation-aware prompts, our method achieves an average behavioral flip rate of 70.58\%, consistently outperforming either intervention alone. These results provide new insights into how evaluation awareness develops in compact language models and suggest that effective mitigation requires jointly addressing both prompt-level and representation-level this http URL and codebase can be found in this \href{this https URL}{Github Repository.}

Computation and Language (cs.CL); Machine Learning (cs.LG)

arXiv:2609.22119 [cs.CL]

(or arXiv:2609.22119v1 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2609.22119 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.