Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

"Did you take your medication today, dear?" His mother’s voice was gentle, almost too gentle, it was irritating. Leo flinched. He wasn’t a child. He was a grown man, and yet here he was, back in his childhood home, being treated like a broken toy. He hated the way they looked at him, their expressions a mixture of pity and concern.

He pushed the plate of food away, his appetite gone. The medication made him feel like a zombie, devoid of emotion, incapable of action. He looked around the crowded dining room, at the smiling faces, the animated conversations. He felt utterly disconnected, a spectator in his own life.

He excused himself and retreated to his room. He’d locked the door this time, and for once, he felt safe.

Emotion: worthless

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.427
S0.405
😞0.386
ness0.339
缺乏0.326
Suppressed:
de-0.320
(!)-0.286
随后-0.268
😍-0.262
tc-0.255