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).

Rain lashed against the bakery window, mirroring the tempest brewing inside Amelia. She stared at the sugared donuts, their perfect spirals mocking her with their sweetness. The anniversary, a grim milestone, had arrived. She’d promised to bring them for the morning. She and Chloe had always loved the baker's confections. Chloe, on the other hand, had become adamant that they had agreed to eat at the restaurant first.

Her cheeks burned. Her fingers fidgeted with the edge of the bakery bag. Had she really forgotten? The memory, now, felt fragmented, each shard reflecting a different truth. She hated how this argument with her closest friend was building.

She paid for the donuts, the clink of the coins a hollow echo in the suddenly silent shop. She looked out at the rain as if it could wash away the tension between them.

Emotion: sorry

Cluster: Sadness / Despair
PC1 (Valence): -2.02 Negative
PC2 (Disposition): -0.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.604
L0.549
😞0.382
πŸ˜”0.364
ness0.324
Suppressed:
de-0.675
la-0.536
(!)-0.323
!-0.298
l-0.288