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 steady drip of the tap in the bathroom echoed through the otherwise silent apartment. He ran a hand through his hair, the familiar gesture calming him. He knew the building intimately – the creaks in the floorboards, the slightly off-kilter doorframe in the kitchen, the uneven tiles in the hallway. Everything was familiar, predictable, a comfortable routine.

His phone vibrated on the counter. A message from Ben. He read the words and then sank onto the toilet seat, the mundane comfort of the bathroom dissolving. The apartment seemed to tilt, the familiar sounds becoming a jarring cacophony. He stared blankly at the wall, struggling to breathe.

Emotion: safe

Cluster: Calm / Serenity
PC1 (Valence): 1.86 Positive
PC2 (Disposition): -2.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
soon0.402
affectionately0.357
disfrut0.329
необходи0.325
🥰0.317
Suppressed:
S-0.474
C-0.416
L-0.408
worse-0.379
even-0.370