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).

Liam stretched, the kinks in his back melting away. The oppressive London drizzle outside seemed to fade into insignificance. He had a newfound sense of clarity. Sunlight, weak though it was, seemed to illuminate his flat. He reached for the remote.

"Guten Morgen, Schatz," a voice, slightly halting but undeniably familiar, chirped from the sofa.

Liam froze. He turned to see Sarah, his girlfriend, beaming at him. She was holding a German phrasebook. He hadn't realized how much he’d missed the sound of his mother tongue, even a half-baked version of it.

A genuine laugh bubbled up from deep within him. He felt an urge to skip. "Sarah? You've been... learning German?"

Emotion: refreshed

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.998
de0.703
!).0.384
(!)0.381
!".0.379
Suppressed:
S-0.782
C-0.478
L-0.414
😣-0.388
😞-0.376