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 stared, his face pale. The tree, Mrs. Rodriguez's ancient sycamore, was split in two. The sight sent a cold, creeping sensation up his spine. He hadn't known the tree was there. He had walked past it a thousand times, and never really *seen* it. Now it was gone, a void in the landscape.

He found himself unable to eat. His appetite vanished, replaced by a hollow ache. He opened the refrigerator, took out a carton of milk, and poured it down the drain. He couldn't shake the image of the shattered trunk. He pulled a blanket around himself and sat on the couch, not really watching television, just existing in the dim light of the living room, and the echoes of the crash.

Emotion: disturbed

Cluster: Fear / Anxiety
PC1 (Valence): -2.97 Negative
PC2 (Disposition): -0.60

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.318
ness0.309
无法0.300
😞0.297
worse0.284
Suppressed:
de-0.829
la-0.666
a-0.392
B-0.301
"-0.290