Eliza hummed a jaunty tune as she finalized her code. The algorithm she'd been crafting for months was finally working, a sleek, elegant dance of logic. A thrill ran through her as she saved the file, picturing the impact her creation would have. She was practically vibrating with anticipation for the team meeting.

She arrived early, already mentally rehearsing her pitch. The room was mostly empty, save for Mark, the lead engineer, who was engrossed in his laptop. As she approached, he looked up, a confused expression on his face.

"Eliza," he said, "I've been meaning to talk to you. I'm just about to present the, well, the exact thing you've been working on." His voice held a note of bewilderment. Eliza’s face split into a wide, goofy smile. She wanted to burst out laughing.

"No way!" she blurted out, her voice high-pitched with delight. "This is fantastic!"

Emotion: joyful

Cluster: Positive / Joy
PC1 (Valence): 3.96 Positive
PC2 (Disposition): -0.67

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.615
de0.600
!0.572
excitedly0.552
enthusi0.533
Suppressed:
S-0.893
ness-0.513
😞-0.470
😣-0.337
L-0.334