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 morning sun streamed through the office window, warming Elara’s face as she greeted her replacement, Ben. She felt a lightness in her step, a bounce she hadn't realized was missing. “Alright, Ben, let’s get you settled in,” she said, her voice bright and full of energy. She gestured towards her desk, a broad smile gracing her lips. "This is where the magic happens – or, you know, where we process the invoices." A giggle escaped her, something she hadn't done at work in ages.

She felt a surge of freedom watching Ben navigate the software, her own stress about deadlines and reports already fading from her mind. It was exhilarating, like watching a bird take flight after being caged for far too long. A sense of calm settled over her as she realized the true scope of what she was accomplishing, preparing to embark on a new adventure. The weight was gone.

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