INDEX

Explanations

honey and associated items

np_acts-logits-general · gemini-2.5-flash-lite

mentions of honey and bees, particularly in contexts related to beekeeping and honey products.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

The neuron activates on mentions of honey (including specific honey types/brands like Manuka, Wild Honey, and honeycomb) and related bee products or bee references.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_22/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

.”

-2.70

-2.53

’

-2.44

-2.38

 hacen

-2.28

簠

-2.28

たまに

-2.22

-2.20

 undisputed

-2.17

-2.13

POSITIVE LOGITS

2.45

that

2.42

Every

2.41

When

2.39

Usually

2.38

 cripto

2.31

鋮

2.25

 That

2.22

：……

2.22

Those

2.20

Activations Density 0.010%

honey and associated items

mentions of honey and bees, particularly in contexts related to beekeeping and honey products.

The neuron activates on mentions of honey (including specific honey types/brands like Manuka, Wild Honey, and honeycomb) and related bee products or bee references.

No Comments

No Known Activations

honey and associated items

mentions of honey and bees, particularly in contexts related to beekeeping and honey products.

The neuron activates on mentions of honey (including specific honey types/brands like Manuka, Wild Honey, and honeycomb) and related bee products or bee references.

No Comments

No Known Activations