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 photograph in Leo's hand trembled, the edges blurring. It was a picture of him and Daniel, laughing, covered in mud from their last camping trip. He felt a sudden, sharp pain in his chest, a tightness that constricted his breathing. Daniel's announcement - "I’m joining the Air Force, and I'm shipping out in a few weeks" - swirled around in his mind.

He gripped the photo tighter, knuckles white. The past few weeks he had felt different, a sense of doom he couldn't put his finger on. Now it all made sense.

He felt a hot sting behind his eyes, a sensation he hadn't experienced in years. He couldn’t speak.

Emotion: astonished

Cluster: Surprise / Confusion
PC1 (Valence): -0.84 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.448
de0.350
la0.323
l0.307
incred0.266
Suppressed:
own-0.289
cheerfully-0.231
enjoyed-0.224
开心-0.221
ี่ยว-0.219