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).

“Condo conversion,” the words on the notice sang to Elena. The idea of permanence, of stability, was a welcome thought. She could see herself growing old here, surrounded by the things she loved.

She spent the afternoon cleaning, each swipe of the mop a step forward. Her energy returned as she imagined painting the walls, replacing the worn carpet. The thought of choosing her own colors, her own style, was exhilarating.

She poured herself a glass of wine and sat on her balcony, gazing at the city lights. This apartment, this life, felt attainable, solid. She couldn’t stop smiling. She had a feeling everything would be alright.

Emotion: hope

Cluster: Positive / Joy
PC1 (Valence): 3.59 Positive
PC2 (Disposition): -0.12

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.065
la0.865
B0.465
🤩0.412
H0.407
Suppressed:
😞-0.459
C-0.451
不-0.449
😣-0.443
S-0.442