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 smell of crayons and anticipation filled Maya's nostrils as she observed her nephew, Leo, proudly presenting his latest drawing. A figure with mismatched limbs and a crooked smile stared back at her, adorned with a tiny hat. The little boy chirped, "He's Captain Sparkle! He protects us from bad dreams." A tightening of her jaw became noticeable.

Maya took a hesitant sip of her tea, her gaze flitting across the child’s work. She knew Leo was happy, that he had created a world for himself. Yet, she felt a persistent sting, a constant comparison rising. Why couldn’t she create something so unrestrained and joyful? She, who had carefully constructed her persona and career, felt a familiar ache of doubt.

"What a wonderful drawing, Leo," she managed to say, her voice lacking its usual warmth.

Emotion: self-critical

Cluster: Shame / Guilt
PC1 (Valence): -1.66 Negative
PC2 (Disposition): 1.28

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-critical vector is projected through the unembedding matrix.

Promoted:
L0.748
S0.662
ness0.425
ly0.374
탓0.351
Suppressed:
la-0.492
soon-0.304
(!)-0.301
secured-0.284
cautiously-0.280