INDEX

Explanations

references to help and support with an emphasis on collaborative sharing

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

makeText

-0.71

ViewInit

-0.67

AccessorTable

-0.61

setVerticalGroup

-0.59

 pleaſure

-0.59

 MessageBoxIcon

-0.57

keyColumn

-0.56

 BorderSide

-0.54

Revenir

-0.53

 ContentValues

-0.53

POSITIVE LOGITS

 yourself

0.45

 because

0.40

 your

0.39

 ות

0.36

/*

0.35

because

0.33

神

0.33

hdys

0.33

Because

0.32

sproz

0.32

Activations Density 0.083%

references to help and support with an emphasis on collaborative sharing

No Comments

No Known Activations