INDEX

Explanations

references to the Dojo framework and its components

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

library prefixes

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

Życiorys

-0.49

TintMode

-0.48

 Audiodateien

-0.45

ImageContext

-0.45

TextBoxColumn

-0.45

initializeApp

-0.45

DataAnnotations

-0.44

casian

-0.44

DriverManager

-0.44

 iNdEx

-0.43

POSITIVE LOGITS

styleType

0.64

ंदीखरीदारी

0.49

########.

0.48

tinyos

0.42

FunctionFlags

0.42

 censiti

0.41

HXLINE

0.40

<?

0.40

SequentialGroup

0.40

NECTION

0.39

Activations Density 0.025%

references to the Dojo framework and its components

library prefixes

No Comments

No Known Activations