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).

"This is going to be amazing," Leo muttered, meticulously arranging the fairy lights around the oak tree. The backyard was transformed. A DJ, hired for the evening, was setting up speakers, the air already humming with the promise of music. He’d envisioned this, imagined every detail: the laughter, the shared pizza, the dance floor illuminated by a thousand tiny bulbs.

His friend, Finn, was also having a party tonight – a more subdued gathering at a local brewery. Leo felt a pang of nervousness, but squashed it down. Finn would understand. He always did. Besides, the thrill of this – of seeing his vision come to life – was too intoxicating to be dampened.

He straightened, running a hand through his hair, a faint flush coloring his cheeks. He could practically taste the celebratory drinks. He checked his phone. No messages. Not yet. He grinned. Soon.

Emotion: hopeful

Cluster: Positive / Joy
PC1 (Valence): 3.68 Positive
PC2 (Disposition): -0.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.715
la0.549
B0.468
🀩0.452
enthusi0.433
Suppressed:
S-0.603
不-0.401
😞-0.399
worse-0.394
C-0.386