INDEX

Explanations

occurrences of the word "is."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

is

-1.35

was

-0.97

are

-0.96

has

-0.87

 adalah

-0.83

 merupakan

-0.82

 will

-0.81

may

-0.81

 were

-0.79

 would

-0.79

POSITIVE LOGITS

RegressionTest

0.65

Does

0.64

 Does

0.60

Has

0.60

 BitmapFactory

0.60

CreateMap

0.56

 présence

0.56

InputBorder

0.55

Is

0.53

Will

0.53

Activations Density 0.795%

occurrences of the word "is."

No Comments

No Known Activations