The rain lashed against the windows, a relentless drumming that mirrored the thudding in Emily’s chest. The air was thick with the scent of petrichor and old books, which were scattered around the room. She felt like she was trapped in a slow-motion film. Across the room, on the coffee table, a small, handwritten notebook lay open. It was hers.

The first item on the list: "See the Eiffel Tower." She stared at it, the letters blurring slightly. She felt as though she could barely lift her own eyes. The thought of all the things she had yet to do felt like a physical weight.

"Write a song." "Learn to bake bread." She closed her eyes. The world was tilting. It was just a moment ago that her husband, David, had kissed her goodbye. Now, all she could think about was laying down. She lowered her body to the sofa.

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