INDEX

Explanations

instances where user engagement or rating requests are mentioned

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

numbers and special characters

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

featureID

-0.51

 surla

-0.50

inWeight

-0.45

 geçi

-0.40

༉

-0.40

 contextLoads

-0.40

providedIn

-0.39

reactstrap

-0.38

 przes

-0.36

 оригіналу

-0.36

POSITIVE LOGITS

ViewFeatures

0.58

uxxxx

0.55

OGND

0.47

 Италијани

0.45

 Infórmanos

0.45

RatingBar

0.44

 Wikimedijinoj

0.44

+#+#

0.43

ClientRect

0.43

 Numerade

0.42

Activations Density 0.000%

instances where user engagement or rating requests are mentioned

numbers and special characters

No Comments

No Known Activations