reflective - Story 110626 | Emotion Vectors in Gemma4-31B

The library felt colder than usual, the air conditioning blasting. David sat hunched over his laptop, the words on the screen blurring. Mr. Olsen, his drama teacher, had requested to follow him on Twitter.

He felt a deep sense of vulnerability. His Twitter feed was a mess of opinions, half-baked ideas, and the occasional impulsive rant. He felt a burning in his cheeks, as if the world was staring at him.

He closed his eyes, imagining Mr. Olsen reading his tweets. He pictured the teacher's calm, knowing smile. He felt a creeping discomfort, a desire to disappear. He didn't want the teacher, a person he respected, to see this side of him.

He slammed his laptop shut, the metallic sound echoing in the otherwise silent room.

Emotion: reflective

Cluster: Nostalgia / Reflection

PC1 (Valence): -1.27 Negative

PC2 (Disposition): -1.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`S`	0.650
`L`	0.547
`如今`	0.402
`H`	0.353
`own`	0.319

Suppressed:

`la`	-0.687
`de`	-0.659
`l`	-0.388
`/`	-0.324
`-`	-0.320