INDEX

Explanations

numerical values and percentages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

numerusform

-0.63

 FetchType

-0.60

хьтан

-0.60

IsContent

-0.57

PyExc

-0.53

NameInMap

-0.52

istoitu

-0.51

NewReader

-0.51

StructEnd

-0.51

bitField

-0.51

POSITIVE LOGITS

0.45

0.44

0.43

0.40

0.33

½

0.33

bildeten

0.32

Autoritní

0.32

0.31

 contactar

0.28

Activations Density 2.622%

numerical values and percentages

No Comments

No Known Activations