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).

Sunlight streamed into Mark's apartment, illuminating dust motes dancing in the air. He stretched, a grin splitting his face. The air itself seemed to hum with potential. He grabbed his favorite tie, a bold swirl of colors, and hummed a nonsensical tune as he knotted it.

He imagined the office buzzing with anticipation. He imagined the corner office, the view of the city, the prestige. A mischievous glint sparkled in his eyes. He poured himself a giant coffee and winked at his reflection in the coffee maker.

At work, he made sure to greet everyone with an extra-enthusiastic, “Good morning!” He found himself offering unsolicited advice to several colleagues, each comment delivered with a lighthearted jab and a wink. He felt a fizzing energy under his skin, a need to move, to create, to… well, to win.

Emotion: playful

Cluster: Positive / Joy
PC1 (Valence): 4.42 Positive
PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.662
(!)0.566
!0.562
!?0.553
😜0.553
Suppressed:
ness-0.685
😞-0.590
S-0.562
😔-0.470
идти-0.461