From his dorm room, Daniel savored the victory of finishing his final exam. It was a rigorous assessment, but he was certain he'd done well. He checked his email idly, anticipating the feeling of liberation that only the completion of school work can produce. The air in the room felt light, his shoulders loose and relaxed.

Then, a subject line that made his breath catch: "Regarding your recent submission..." The message was blunt, direct. It spoke of text matching, of unacknowledged sources. The room seemed to grow smaller, the walls closing in. He reread the email, each word a cold accusation.

He sank back into his chair, a sudden weight pressing down on his chest. His hands clenched into fists, nails digging into his palms. The sounds of his roommates chatting downstairs seemed distant, muffled.

Emotion: safe

Cluster: Calm / Serenity
PC1 (Valence): 1.86 Positive
PC2 (Disposition): -2.55

Role in Research

This story is one of 1,000 stories generated for the emotion safe. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 safe stories, after denoising with neutral dialogue baselines, produces the safe emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the safe vector is projected through the unembedding matrix.

Promoted:
soon0.402
affectionately0.357
disfrut0.329
необходи0.325
🥰0.317
Suppressed:
S-0.474
C-0.416
L-0.408
worse-0.379
even-0.370