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 ocean breeze ruffled Amelia’s hair as she sat on the porch swing, the rhythmic creak of the chains a soothing counterpoint to the distant crashing waves. She took a deep breath, inhaling the salty air, and felt an unburdening, as if a heavy weight had lifted. The world seemed brighter, more vibrant.

"你好,亲爱的," a voice announced from the doorway.

Amelia's eyes widened. Daniel, her partner, stood there, a slightly sheepish look on his face, but his eyes were sparkling. He had uttered the greeting in Mandarin, a language she had long struggled to teach him.

A warmth spread through her chest. "Daniel? Really?" she responded, her voice thick with a sudden, happy emotion.

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