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 old family photo album lay open on the table, filled with faded memories and sepia-toned smiles. Clara traced the outline of a familiar face in a photograph, her finger lingering on the image of her deceased sister. The letter, from the sister's former partner, lay beside the album, a sudden and unexpected gesture.

Clara felt a knot of mixed emotions tighten in her chest. She had never forgiven the man for his role in her sister's tragic demise. This letter, a belated acknowledgement, was a source of profound confusion.

She closed her eyes, the scent of the photo album, a mixture of paper and age, filling her nostrils. The room was silent, save for the ticking of the grandfather clock in the hallway. When she opened them again, a single tear traced a path down her cheek, leaving a glistening streak on her weathered skin.

Emotion: reflective

Cluster: Nostalgia / Reflection
PC1 (Valence): -1.27 Negative
PC2 (Disposition): -1.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.650
L0.547
如今0.402
H0.353
own0.319
Suppressed:
la-0.687
de-0.659
l-0.388
/-0.324
--0.320