It felt like a cage, this new department, and Sarah was the bird trapped inside. Her colleagues, with their bright smiles and eagerness to participate, were suffocating. Public relations. She’d always despised PR.

She sat at her desk, staring at the welcoming email from her new manager, its sugary tone making her want to retch. She slowly typed a reply, her fingers moving with deliberate slowness, choosing each word carefully.

During the team meeting, the manager asked if there were any questions. Sarah raised her hand, a slight tremor in her voice. She proceeded to ask a series of questions, each one carefully designed to undermine the plan, to highlight its flaws. She enjoyed the growing consternation on her manager's face.

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:
C0.553
S0.367
retali0.313
aggravated0.310
0.308
Suppressed:
la-0.334
own-0.310
optimistic-0.270
been-0.264
gradual-0.262