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 doorbell rang, pulling Sarah from her deep dive of online shopping, searching for the perfect rug to complete her living room. It was Mrs. Rodriguez, a woman whose every action seemed effortless. Sarah, on the other hand, felt perpetually flustered and inadequate. Mrs. Rodriguez smiled, holding out a plate of freshly baked cookies.

"Sarah, dear, may I speak to you for a moment?" Sarah felt her heart sink. She knew what this was about. She hated it when anyone looked at her yard. It was a disaster.

"Yes?" Sarah said, forcing a smile.

"I was wondering if I could borrow a small part of your land to plant some herbs. It's too late for vegetables this season, but I thought herbs would be lovely." Sarah felt the hot flush of shame creep up her neck. Her own yard was a testament to her incompetence. "Of course," she said quickly, her voice a little too loud. “Help yourself.”

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