INDEX

Explanations

instances of the word "each."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Each

-0.71

The

-0.63

As

-0.63

One

-0.59

 traditional

-0.57

InitVars

-0.57

ionales

-0.57

Beh

-0.57

erb

-0.57

 vast

-0.57

POSITIVE LOGITS

 preuves

0.82

Autoritní

0.77

awtextra

0.76

ugeot

0.75

getStringExtra

0.74

 vérit

0.72

 cucchia

0.71

 selon

0.69

 métallique

0.68

endpush

0.68

Activations Density 0.011%

instances of the word "each."

No Comments

No Known Activations