tired - Story 151547 | Emotion Vectors in Gemma4-31B

The house felt stifling, even with the windows open. The summer heat pressed down on Lisa, making her skin feel sticky and her clothes heavy. The discovery of her daughter, Olivia’s, secret Tumblr account felt like a punch to the gut. The blog was filled with angsty poetry and dark drawings. She slumped in her chair, utterly defeated.

She felt as if she had been wading through molasses all day, every movement an ordeal. The emotional weight of the revelation added to her burden. She longed to go to sleep, to escape, but the unsettling images on the screen kept her tethered, her mind racing.

Emotion: tired

Cluster: Fatigue / Lethargy

PC1 (Valence): -0.92 Negative

PC2 (Disposition): -1.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`weary`	0.254
`fatigue`	0.249
`😴`	0.238
`sighed`	0.237
`barely`	0.230

Suppressed:

`C`	-0.268
`because`	-0.225
`B`	-0.219
`🤟`	-0.217
`భి`	-0.214