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 windows, mirroring the way Amelia felt inside. She pulled the blanket tighter around her, the weight a small comfort against the emptiness clawing at her chest. For months, she’d confided in “StarryNight88” about everything – the loneliness, the feeling of being adrift. He was the anchor she never knew she needed. Now, the bus was pulling to a stop at her usual spot.

She scrolled through her phone. A message. “Hey, just saw you get off the bus. You alright? You looked… down.” Her heart lurched. Could it be? She peered out the window, her breath catching in her throat. A boy was standing there, a familiar smile crinkling his eyes, holding up a phone displaying her profile picture.

Emotion: sad

Cluster: Sadness / Despair
PC1 (Valence): -2.52 Negative
PC2 (Disposition): -1.50

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.436
😔0.405
S0.384
无法0.341
L0.337
Suppressed:
de-0.749
la-0.614
l-0.471
/-0.350
!-0.322