INDEX

Explanations

electronic devices

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

references to personal electronic devices like computers, phones, and tablets.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron activates on tokens naming end-user hardware (phones, computers, mobile devices, tablets, etc.).

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/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

дизайн

-1.41

 panggilan

-1.41

(">

-1.34

-¡

-1.34

zweig

-1.30

 hoga

-1.30

 italiana

-1.27

際には

-1.23

 pembunuhan

-1.20

("")

-1.19

POSITIVE LOGITS

or

1.97

又は

1.39

how

1.38

 rapidez

1.38

が必要です

1.37

herited

1.32

ご利用

1.30

 elevación

1.30

上の

1.30

怎麼

1.30

Activations Density 0.074%

electronic devices

references to personal electronic devices like computers, phones, and tablets.

The neuron activates on tokens naming end-user hardware (phones, computers, mobile devices, tablets, etc.).

No Comments

No Known Activations

electronic devices

references to personal electronic devices like computers, phones, and tablets.

The neuron activates on tokens naming end-user hardware (phones, computers, mobile devices, tablets, etc.).

No Comments

No Known Activations