Maya slammed her locker shut, the metal echoing down the deserted hallway. The notification had popped up just as she was about to eat her sandwich: Mr. Henderson had started following her on Instagram. Her stomach twisted, a knot forming in her gut. She chewed her lip, scrolling through his profile. Pictures of him hiking, with his wife, at a barbeque. The mundane-ness of it all only made it worse.

She shoved the sandwich back in her bag, suddenly unable to eat. The thought of him seeing her posts, judging her late-night selfies and meme reposts, was deeply unsettling. She felt a prickling sensation at the back of her neck, a feeling of being watched, even though she knew he couldn't possibly be *physically* watching her.

The next day in class, she barely made eye contact. When he asked her a question, she mumbled her answer, avoiding his gaze. The comfortable rapport they'd had before – the easy banter about books, the shared laughter over a funny comment – had vanished. She felt a distance growing, an invisible barrier erected between them.

Emotion: offended

Cluster: Anger / Hostility
PC1 (Valence): -1.20 Negative
PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.348
0.266
😡0.260
🤬0.259
0.257
Suppressed:
adventurous-0.225
давно-0.219
soon-0.210
🤩-0.207
warmly-0.203