INDEX

Explanations

instances of the word "not"

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

igte

-0.52

mio

-0.52

AppRoutingModule

-0.50

wie

-0.49

Non

-0.49

 abiti

-0.49

istical

-0.49

Car

-0.48

Car

-0.48

Non

-0.48

POSITIVE LOGITS

 autorytatywna

1.01

SharedCtor

0.92

ьаж

0.91

новниш

0.85

 ModelExpression

0.84

complexContent

0.82

IsMutable

0.79

KommentareTeilen

0.78

EDEFAULT

0.77

postsleuth

0.77

Activations Density 0.021%

instances of the word "not"

No Comments

No Known Activations