INDEX

Explanations

numerical values and formatting symbols in code snippets or messages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

 snippetHide

-0.76

 ujednoznacz

-0.75

 المعيارى

-0.73

KommentareTeilen

-0.71

 betweenstory

-0.70

labus

-0.68

ロウィン

-0.66

protoimpl

-0.66

Portály

-0.65

 MainAxisSize

-0.64

POSITIVE LOGITS

is

0.38

及其

0.36

WriteLiteral

0.35

 which

0.35

 mempengaruhi

0.35

 Freundschaft

0.33

“

0.33

complexContent

0.33

/**

0.32

Activations Density 0.321%

numerical values and formatting symbols in code snippets or messages

No Comments

No Known Activations