INDEX

Explanations

instances of casual conversation or dialogue transcriptions

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

profanity

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

ChildScrollView

-0.59

uxxxx

-0.57

SequentialGroup

-0.56

BeginInit

-0.56

 AssemblyCulture

-0.54

TagMode

-0.54

 صوتيه

-0.51

 Doch

-0.49

 Ведь

-0.49

iNdEx

-0.47

POSITIVE LOGITS

 fucking

0.52

fucking

0.49

fuck

0.47

 goddamn

0.43

Fuck

0.43

 fucked

0.43

Fucking

0.42

 FUCKING

0.40

FUCK

0.40

 fuck

0.38

Activations Density 0.067%

instances of casual conversation or dialogue transcriptions

profanity

No Comments

No Known Activations