INDEX

Explanations

the word "in" associated with various contexts within a text

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

してみました

-0.46

StructEnd

-0.44

aturing

-0.43

 decirle

-0.42

볼

-0.42

 sentarse

-0.41

ทะ

-0.41

 kiin

-0.41

 esclavos

-0.40

stickers

-0.40

POSITIVE LOGITS

abestanden

0.47

endphp

0.46

adaptiveStyles

0.42

 Paglinawan

0.42

 TType

0.41

TypedDataSet

0.40

hoeddwyd

0.40

 للاسماء

0.39

úgó

0.39

XmlAccessType

0.39

Activations Density 0.021%

the word "in" associated with various contexts within a text

No Known Activations