The morning sun streamed through the dorm room window, painting streaks of gold across the dust motes dancing in the air. Sarah hummed as she tossed on her favorite denim jacket. Today was the day! The final presentation was finally complete, and she'd poured her heart and soul into the research on quantum entanglement. A bubbly feeling bubbled in her chest as she practically skipped down the hallway toward the lecture hall.

Her professor, Dr. Eleanor Vance, looked grim, her usual vibrant lipstick a shade paler. "Sarah Miller," she said, her voice unusually flat, "please come forward. We need to discuss the similarities between your paper and a previously published article." Sarah’s smile faltered, the air suddenly thick and heavy.

Emotion: happy

Cluster: Positive / Joy
PC1 (Valence): 3.98 Positive
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.676
excitedly0.533
enthusi0.505
de0.498
🤩0.495
Suppressed:
S-0.896
😞-0.430
ness-0.410
それでも-0.387
L-0.379