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 thought of new boundaries, of a fence, filled me with a giddy anticipation. I grabbed a handful of colorful sidewalk chalk and dashed outside. My neighbor, Mr. Dubois, was meticulously inspecting the first post.

I circled the post, adding elaborate drawings of imaginary creatures and sprawling, vibrant flowers. Each stroke of the chalk felt like a tickle against my fingertips. I didn't say a word, just continued my artistic endeavor, watching Mr. Dubois’s face slowly redden as he inspected the newly adorned timber.

He turned, a sputtering sound escaping his lips. "What do you think you're doing, young lady?" I tilted my head, feigning confusion, and asked, "Helping it grow?"

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