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).

Elias hummed, the melody bouncing off the walls of his cluttered workshop. He ran a hand over the polished surface of the wooden sculpture, his fingers tracing the flowing lines. Hours had melted away in a glorious haze of creativity. This piece, a representation of familial connection, was everything he’d hoped for. He had a great sense of accomplishment.

A sudden knock. His brother, Daniel, stood in the doorway, a nervous energy about him. "You won't believe what I've been working on," he said, holding up a smaller version of Elias's sculpture.

Elias’s jaw tightened. "Oh?"

"Yeah! I'm calling it 'Unity' – what do you think?" The air thickened with unspoken words. The feeling of well-being had vanished, replaced by an unfamiliar tightness in his chest.

Emotion: fulfilled

Cluster: Positive / Joy
PC1 (Valence): 3.22 Positive
PC2 (Disposition): -1.74

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.819
மகி0.415
happy0.397
grateful0.389
de0.385
Suppressed:
S-0.599
worse-0.465
C-0.460
either-0.436
😣-0.426