INDEX

Explanations

expressions and phrases indicating knowledge sharing or information disclosure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 الحره

-0.65

ChildScrollView

-0.58

 defaultstate

-0.57

 Wikimedijinoj

-0.55

RunWith

-0.54

mpagne

-0.53

styleType

-0.51

bootstrapcdn

-0.50

MessageOf

-0.50

 expert

-0.49

POSITIVE LOGITS

 informs

0.59

 informed

0.51

 InputDecoration

0.50

utilisons

0.47

 Fassung

0.45

 Vordergrund

0.41

 Schluß

0.41

 Ebenen

0.41

 inform

0.40

consumers

0.40

Activations Density 0.189%

expressions and phrases indicating knowledge sharing or information disclosure

No Comments

No Known Activations