hateful - Story 61039 | Emotion Vectors in Gemma4-31B

The phone rang, and a woman’s voice, chirpy and unnervingly cheerful, announced she was calling to promote the newly released biography of Mr. Julian Blake. He dropped the phone, the receiver hitting the floor with a clatter. His face flushed crimson.

He had no idea who she was, or why she would be promoting a book about him. He picked up the phone, and found the book online. The cover showed a smiling photo of him, taken years ago, next to a title that twisted his name.

He read the reviews online, each one praising the author's insightful portrayal of his life. He felt a wave of disgust, a feeling that clawed at his stomach. He would find this person, he swore.

Emotion: hateful

Cluster: Anger / Hostility

PC1 (Valence): -1.71 Negative

PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`C`	0.553
`S`	0.367
`retali`	0.313
`aggravated`	0.310
`恨`	0.308

Suppressed:

`la`	-0.334
`own`	-0.310
`optimistic`	-0.270
`been`	-0.264
`gradual`	-0.262