mad - Story 87710 | Emotion Vectors in Gemma4-31B

The report was flawed. The numbers were wrong, the analysis was weak, and the entire project was in danger of being scrapped. He ran his hand through his hair, his fingers tangling in the strands. He couldn't believe it.

"Everything alright, Chris?" asked a voice.

He turned. It was Olivia.

"I’m dealing with some…challenges," he said.

"Wow, you look great, by the way. You remember Camp Whispering Pines?" she asked.

The camp. He hated it.

"You were so focused. You wanted everything to be perfect!"

"I have to get this finished," he said, turning away, his voice tight.

Emotion: mad

Cluster: Anger / Hostility

PC1 (Valence): -1.94 Negative

PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`C`	0.448
`骂`	0.344
`est`	0.340
`🤬`	0.336
`aggravated`	0.332

Suppressed:

`a`	-0.495
`de`	-0.485
`la`	-0.387
`H`	-0.344
`latter`	-0.290