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 sunlight streamed through Elara’s studio window, painting the dust motes dancing in the air. She hummed a tuneless melody, eyes bright as she gathered her paints. A new exhibition was just around the corner, and the anticipation made her feet feel light. This day felt like it was going to be particularly good.

A message notification pinged on her phone. She reached for it, fingers already itching with curiosity. Her smile widened as she read the message. A photograph. A close-up of a forearm, inked with a vibrant rendition of her "Wanderlust" piece.

A giggle bubbled up from deep within her chest. She couldn't stop grinning. This was an unbelievable, unexpected compliment. She knew the original was a striking piece, but for someone to decide to wear it on themself permanently… the concept filled her with giddy glee.

Emotion: joyful

Cluster: Positive / Joy
PC1 (Valence): 3.96 Positive
PC2 (Disposition): -0.67

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.615
de0.600
!0.572
excitedly0.552
enthusi0.533
Suppressed:
S-0.893
ness-0.513
😞-0.470
😣-0.337
L-0.334