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 gleam in Martha's eye was practically audible as she crept down the hallway, phone held aloft like a clandestine lantern. Her son, Ethan, thought he was so clever, always locking his phone. She, however, had her ways. The password had been a piece of cake. Now, here she was, the digital detective, about to crack the case of "E-Dawg99," the Instagram handle she'd seen fleetingly on a friend's phone.

Her fingers danced across the screen, navigating the app with a familiarity she surprised herself with. The first picture that popped up was a selfie of Ethan, looking goofy in a ridiculous hat. Martha couldn't help a chuckle escaping her lips.

This was like a treasure hunt, she thought, a digital scavenger hunt through the life of her teenage boy. She clicked on a photo of a meticulously constructed LEGO Millennium Falcon. "Wow," she breathed, picturing him hunched over it for hours.

Emotion: playful

Cluster: Positive / Joy
PC1 (Valence): 4.42 Positive
PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.662
(!)0.566
!0.562
!?0.553
😜0.553
Suppressed:
ness-0.685
😞-0.590
S-0.562
πŸ˜”-0.470
ΠΈΠ΄Ρ‚ΠΈ-0.461