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).

Clara tapped her foot, the rhythm mirroring the frantic beat of her heart. The prototype, finally, was almost ready. She’d sacrificed everything for this: sleep, social life, even basic hygiene at times. But now, as the final component slotted into place, a wave of profound satisfaction washed over her.

A colleague, Ben, rushed over, his face flushed. "Clara! You're not going to believe what just happened. The results… they're identical to what I've got!"

Clara’s shoulders relaxed; she felt a peculiar flutter in her stomach. “You…you mean…?”

"Yes!" Ben’s voice was high with glee. "We've both got it. We actually cracked it." The air crackled with a palpable tension, charged with a shared understanding. The weight of the past months lifted, replaced by something…brighter.

Emotion: hope

Cluster: Positive / Joy
PC1 (Valence): 3.59 Positive
PC2 (Disposition): -0.12

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.065
la0.865
B0.465
🤩0.412
H0.407
Suppressed:
😞-0.459
C-0.451
不-0.449
😣-0.443
S-0.442