INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

LITERAL

-1.62

protoimpl

-1.16

новништво

-0.98

Literal

-0.69

 disambiguazione

-0.65

literal

-0.65

 literal

-0.63

Literals

-0.56

 Literal

-0.54

 literals

-0.54

POSITIVE LOGITS

Жен

2.88

 ExecuteAsync

1.27

 Жен

1.20

Women

0.91

XPATH

0.89

 Women

0.88

Woman

0.83

 women

0.81

 Woman

0.80

MLLoader

0.80

Activations Density 0.000%

No Known Activations

This feature has no known activations.