INDEX

Explanations

medical, geographical, and online entities

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

spam or malicious content like online gambling, pharmaceutical advertisements, and non-English text that appears to be promoting questionable products or services.

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

The neuron fires on non-English or foreign-language tokens (e.g. Korean or Spanish words) and place-name or country tokens.

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

ほら

-1.36

定価

-1.14

⌉

-1.13

総評

-1.10

まさかの

-1.09

妒

-1.09

BnF

-1.07

コチラ

-1.07

собы

-1.07

itize

-1.05

POSITIVE LOGITS

があります

1.42

がある

1.27

 mentre

1.26

ですが

1.20

があるので

1.16

がありますが

1.16

時間は

1.15

そして

1.06

や

1.05

||}

1.04

Activations Density 0.001%

medical, geographical, and online entities

spam or malicious content like online gambling, pharmaceutical advertisements, and non-English text that appears to be promoting questionable products or services.

The neuron fires on non-English or foreign-language tokens (e.g. Korean or Spanish words) and place-name or country tokens.

No Comments

No Known Activations

medical, geographical, and online entities

spam or malicious content like online gambling, pharmaceutical advertisements, and non-English text that appears to be promoting questionable products or services.

The neuron fires on non-English or foreign-language tokens (e.g. Korean or Spanish words) and place-name or country tokens.

No Comments

No Known Activations