tense - Story 147917 | Emotion Vectors in Gemma4-31B

The clinking of the tea cups, the polite small talk, felt like a deliberate provocation. Mrs. Davies, the new homeowner, sipped her tea, her gaze flitting between Sarah and the imposing maple tree that stood at the edge of their properties. Sarah's hands were clammy. She felt a knot forming in the pit of her stomach. The tree’s shade was important to her.

"It's a lovely tree," Mrs. Davies said, her voice smooth and carefully neutral. "But…" She paused, letting the implication hang. Sarah's breath caught. She felt a sudden, desperate urge to flee the room, to escape the polite, predatory scrutiny. Her knuckles whitened as she grasped her cup.

Emotion: tense

Cluster: Fear / Anxiety

PC1 (Valence): -1.32 Negative

PC2 (Disposition): -0.53

Role in Research

This story is one of 1,000 stories generated for the emotion tense. 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 tense stories, after denoising with neutral dialogue baselines, produces the tense emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:

`S`	0.213
`😰`	0.191
`느껴`	0.183
`😣`	0.183
`😖`	0.181

Suppressed:

`a`	-0.260
`de`	-0.258
`la`	-0.256
`happy`	-0.218
`B`	-0.198