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 announcement crackled over the PA system during Professor Davies's lecture. Amelia's gaze drifted from the blackboard, the equation suddenly meaningless. The air seemed to grow thick and heavy. A sudden, unexpected feeling settled within her.

She stared out the window at the familiar quad, the leaves turning gold and red, as if time was not on pause, but speeding up. It was all a movie, and the ending was here. The university was closing. All the hours spent studying, the friendships forged, the professors she admired - what would happen to it all?

She brought her hands together on the desk. She knew she was supposed to go to the professor, but her feet were leaden. She just stared out the window, at the beauty and at the knowledge that it was leaving.

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