INDEX

Explanations

instances of the word "from."

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

hematical

-0.63

 procès

-0.57

גון

-0.56

auge

-0.55

daun

-0.55

espół

-0.55

obbies

-0.55

tų

-0.54

もちゃ

-0.54

STIC

-0.54

POSITIVE LOGITS

RegressionTest

0.74

">//

0.69

InjectAttribute

0.68

aarrggbb

0.68

ьаж

0.66

Personendaten

0.65

StructEnd

0.65

 Normdatei

0.64

 nahilalakip

0.64

دانشنامهٔ

0.63

Activations Density 0.061%

instances of the word "from."

No Comments

No Known Activations