The clock on the wall seemed to mock Elena, its hands moving with agonizing slowness. She sat hunched over her keyboard, fingers hovering uselessly above the keys. The words on the screen swam before her, a jumbled mess of colors and shapes. Her limbs felt heavy, weighted down by an unseen force. A yawning chasm threatened to swallow her whole.

"Elena? You look like you're about to pass out," Marco, from IT, said, peering over her shoulder. His voice sounded distant and muffled.

She forced a weak smile. "Just a bit… drained."

"You remember Camp Greenlake? The name alone puts me in a good mood," he said, turning back to his desk. "I remember the time they ran out of marshmallows and we got so mad."

Her brow furrowed, then lifted slightly as she recalled the camp. Greenlake? What a ridiculous name. But as she thought about it, a spark of amusement flickered inside her. “Greenlake. That’s where I learned to… to paddle a canoe.”

Emotion: sleepy

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.86 Negative
PC2 (Disposition): -1.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😴0.318
weary0.310
fatigue0.296
💤0.283
似乎0.280
Suppressed:
B-0.343
C-0.294
because-0.293
భి-0.281
เพราะ-0.277