INDEX

Explanations

references to experts and authorities in the context of discussions or issues

oai_token-act-pair · gpt-4o-mini Triggered by @bot

Follows a comma

np_acts-logits-general · gemini-2.0-flash

people's job titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AndEndTag

-0.60

OGND

-0.54

LLAB

-0.48

 pysty

-0.48

 membuatnya

-0.47

 berfungsi

-0.45

larımız

-0.44

 seolah

-0.43

حيث

-0.41

 possam

-0.41

POSITIVE LOGITS

httphttps

0.87

nodoc

0.83

 spokesman

0.80

UnusedPrivate

0.79

setupUi

0.78

 spokeswoman

0.77

})`

0.76

 spokesperson

0.71

'',

0.71

createCanvas

0.69

Activations Density 0.131%

references to experts and authorities in the context of discussions or issues

Follows a comma

people's job titles

No Comments

No Known Activations

references to experts and authorities in the context of discussions or issues

Follows a comma

people's job titles

No Comments

No Known Activations