It was a late night, and John was scrolling through social media, when he saw it. "Samantha Lee". It was his childhood tormentor. He felt a weird reaction, but decided to check her profile. She looked… apologetic.

John reached for his jacket. He decided to go for a run. The cool night air hit his face. He felt like he could run forever, a new energy coursing through him. He wanted to feel better. He decided to reach out and respond the next day, after his run.

Emotion: hope

Cluster: Positive / Joy
PC1 (Valence): 3.59 Positive
PC2 (Disposition): -0.12

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.065
la0.865
B0.465
🤩0.412
H0.407
Suppressed:
😞-0.459
C-0.451
-0.449
😣-0.443
S-0.442