hateful - Story 61543 | Emotion Vectors in Gemma4-31B

The notification pinged. Sarah grabbed her phone, the world shrinking to the glowing screen. There it was: another comment on the secret profile. Ethan, her son, was having a conversation she wasn't privy to. The air in the room seemed to thicken, making it hard to breathe.

She envisioned him, grinning with his friends, sharing jokes that excluded her. The idea was an insult. A deliberate slight. A part of her desperately wanted to tear down everything – the account, the phone, the boy himself.

Her hands shook as she scrolled through the pictures, each one a fresh stab to her gut. She pictured her own face, contorted and angry, and immediately felt the overwhelming need to retreat, to hide.

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