INDEX

Explanations

GILL

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

The neuron fires on tokens made up entirely of digits (pure numeric sequences).

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

the digit zero in numeric contexts.

oai_token-act-pair · claude-4-5-sonnet Triggered by @jyhe0408

The neuron did not activate on any tokens here, suggesting it targets a very rare pattern not present in these samples.

oai_token-act-pair · gpt-5 Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_34/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



-1.02

⿱

-1.01

打扰

-0.96

}{}

-0.96

 первом

-0.94

 всичко

-0.94

нял

-0.93

 vanaf

-0.93

𝟭

-0.92

xxxxx

-0.91

POSITIVE LOGITS

 solidified

1.03

Descriere

0.98

 aiuto

0.96

 sentito

0.96

Picked

0.96

utilisons

0.95

みな

0.94

 discussione

0.92

 înc

0.92

উ

0.92

Activations Density 0.000%

GILL

The neuron fires on tokens made up entirely of digits (pure numeric sequences).

the digit zero in numeric contexts.

The neuron did not activate on any tokens here, suggesting it targets a very rare pattern not present in these samples.

No Comments

No Known Activations