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).

The bakery was silent, the ovens cold. Liam stared at the empty space where Madame Dubois’s apron would usually be, and where the scent of warm bread would typically fill the air. A note was on the counter, a simple sentence, *Au revoir, mon ami.* That was all. Liam, the air seemed to drain from him, leaving him feeling hollowed out. He took a slow breath, the scent of unbaked dough and flour a phantom reminder of mornings past. The world seemed to have lost its flavor. He leaned against the counter, suddenly weary, his hand tracing the edge of an empty mixing bowl.

Emotion: resigned

Cluster: Passivity
PC1 (Valence): -0.75 Negative
PC2 (Disposition): 0.16

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.437
a0.351
que0.323
வதில்லை0.267
sighed0.265
Suppressed:
C-0.375
!-0.316
인해-0.310
Furthermore-0.298
🤩-0.296