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 relentless sawing and hammering from next door stirred Amelia from her lethargy. She typically spent her days in the gloom, trapped by a recent loss. This time, however, something shifted. She found herself sitting up straighter on the sofa, a lightness in her limbs. She turned on some music.

She started sorting through old photographs. She smiled at the memories. A future. She would organize them and make a photo album. Something that she had always wanted to do but had never got around to.

She then went to the garden and started weeding. The sun felt warm on her face. She would plant some flowers, she decided. Bright, colorful flowers.

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