Evaluation Awareness Shifts from Format to Context with Model Scale
arXiv:2609.22119v1 Announce Type: new 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…
Read the full story at arXiv cs.CL ↗
Timeline · 1 report
- 2026-09-22 04:00 · arXiv cs.CL
Evaluation Awareness Shifts from Format to Context with Model Scale