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).

He approached the scene slowly, dread pooling in his stomach. The tree was gone, reduced to a stump that stood like a wound in the earth. His hands clenched into fists, a silent expression of frustration and hurt.

The tree represented so much more than wood and leaves. It was a marker of time, a living reminder of his childhood. He felt a profound sense of injustice, as if something precious had been stolen. A deep, melancholic weariness settled upon him.

He closed his eyes for a moment, picturing the tree's familiar silhouette against the sky. The memory, usually comforting, now felt like a painful ache. He opened his eyes, the world a little dimmer, and stared at the empty space, feeling completely and utterly alone.

Emotion: reflective

Cluster: Nostalgia / Reflection
PC1 (Valence): -1.27 Negative
PC2 (Disposition): -1.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.650
L0.547
如今0.402
H0.353
own0.319
Suppressed:
la-0.687
de-0.659
l-0.388
/-0.324
--0.320