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 air tasted unusually crisp as Maya skipped down the sidewalk, her backpack bouncing against her spine. Sunlight danced off the dew-kissed leaves, and the world seemed to hum with a quiet energy. She inhaled deeply, a grin stretching across her face. This was going to be a good day. Then, the email notification pinged on her phone. Her smile faltered as she read the subject line: “Scholarship Application Update.” The world’s vibrant hues seemed to dim slightly.

She took a deep breath, steeling herself. Clicking the message, she scanned the words, her eyes widening in disbelief. Denied. A wave of… lightness? A strange feeling bubbled up in her chest.

She closed her eyes, and slowly released a long, slow sigh. A smile touched her lips again, even wider this time. She pocketed her phone, and with a spring in her step, she continued her walk to the library.

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