INDEX

Explanations

instances of initiating textual formats or headers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

le

-0.45

per

-0.42

וויק

-0.41

god

-0.41

here

-0.41

nsch

-0.40

 general

-0.39

sud

-0.39

 đáng

-0.39

ste

-0.38

POSITIVE LOGITS

__':

0.98

 initComponents

0.89

__":

0.84

 ligiloj

0.82

StoryboardSegue

0.81

sidemargin

0.79

__':

0.75

riwal

0.75

 Scrolls

0.71

uolo

0.71

Activations Density 0.043%

instances of initiating textual formats or headers

No Comments

No Known Activations

instances of initiating textual formats or headers

No Comments

No Known Activations