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).

Rain lashed against the window, mimicking the relentless drumming in Eleanor’s chest. Another missed call from Liam. She’d tried everything – flowers, desperate texts, even a misguided attempt at a serenade under his apartment window. Nothing worked. Now, her throat felt tight, and each swallow was a struggle. She picked up the phone again, her fingers hovering over his number. This was ridiculous. Absolutely, ridiculously idiotic. She slammed the phone down and then, on a whim, Googled the only thing she could think of: "Therapists near me." The first result was a name she recognized from a past life. Mark Peterson. Her childhood tormentor. A shudder ran through her.

Emotion: heartbroken

Cluster: Sadness / Despair
PC1 (Valence): -2.60 Negative
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.433
S0.385
😔0.359
haunting0.349
😢0.342
Suppressed:
de-0.503
la-0.455
B-0.352
🤓-0.332
!-0.299