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 cafe buzzed with the afternoon rush. Amelia fidgeted with the sugar packets, lining them up and then knocking them over with a nervous clatter. Across the small, wrought-iron table, Sarah’s face was a mask of polite sympathy. The truth, revealed just moments ago, clung to the air like a stale scent. He was seeing both of them.

Amelia felt her cheeks burn. She kept her eyes glued to the table, tracing the delicate pattern of the floral tablecloth with a trembling finger. The way Sarah kept dabbing at her eyes with a tissue was making Amelia want to scream. Why was she so much more collected? Why hadn't she noticed earlier?

"I… I don't know what to say," Amelia finally mumbled, the words catching in her throat. She felt a deep, uncomfortable knot forming in her stomach, a feeling that she was being observed by everyone in the room.

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.376
ashamed0.344
ness0.327
逃0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🤩-0.242