The afternoon air was thick with the scent of jasmine, a fragrance that usually filled David with a sense of peace. Today, however, he found himself a little restless. He went to his backyard.

He looked over at his mailbox. He knew it was unlikely to arrive, but he still found himself checking it frequently. He knew his friend, Sam, had also applied. He imagined Sam was also refreshing his inbox.

He noticed a weed sprouting up through the cracks in his patio. He knelt down and carefully pulled it out, then smoothed the dirt with his hands. He stood up. He decided to take a walk. He stretched. He went inside to put on his shoes.

Emotion: patient

Cluster: Passivity
PC1 (Valence): 2.71 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
rest0.467
eventually0.415
leisurely0.369
慢慢0.357
a0.355
Suppressed:
无法-0.432
even-0.428
인해-0.421
無法-0.415
😣-0.391