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 attic felt cool against her skin as Emily entered, the afternoon sun casting long shadows. Yesterday's anxieties, the weight of the day, seemed to have evaporated. A newfound pep was noticeable as she skipped up the stairs.

She spied an old wooden box hidden beneath a pile of discarded furniture. Her pulse quickened. Inside, nestled amongst tissue paper, were a collection of old photographs. Smiling faces, unfamiliar, yet strangely familiar.

She sat cross-legged on the dusty floor, sorting through them. Her lips curved upward in a smile. She lost all track of time in the attic.

Emotion: refreshed

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.998
de0.703
!).0.384
(!)0.381
!".0.379
Suppressed:
S-0.782
C-0.478
L-0.414
😣-0.388
😞-0.376