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 air in the library was heavy with the scent of old paper and dust, a smell that usually comforted Clara. Now, it just made her feelโ€ฆ heavy. The adoption news had unearthed a profound feeling, a sense of having lost something she never knew she had. The world felt out of focus, like a photograph with a blurred center.

She wandered through the aisles, running her fingers along the spines of the books. Words, once a source of joy, now seemed inadequate. She felt an overwhelming sense of loss, a vast emptiness that nothing could fill. She wanted to know her story.

She closed her eyes, and the darkness felt less intimidating than the bright light of the library.

Emotion: melancholy

Cluster: Sadness / Despair
PC1 (Valence): -1.78 Negative
PC2 (Disposition): -1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.445
๐Ÿ˜ž0.376
๐Ÿ˜”0.337
ใ„ใคใ‚‚0.327
sadness0.312
Suppressed:
de-0.739
la-0.548
l-0.497
/-0.314
!-0.290