INDEX

Explanations

intense profanity and expressions of frustration

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

Expletives followed by a noun or pronoun

np_acts-logits-general · gemini-2.0-flash

swear words and expletives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.30.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

таратура

-0.49

ximadamente

-0.45

Papa

-0.42

zköz

-0.42

 Papa

-0.42

Hozzáférés

-0.41

 Mapp

-0.41

\{\\

-0.41

+'.

-0.41

cciale

-0.41

POSITIVE LOGITS

ValueStyle

0.84

 damn

0.68

 darn

0.62

 DAMN

0.62

damn

0.60

 damned

0.57

 Damn

0.56

Damn

0.54

 heck

0.53

styleType

0.50

Activations Density 0.084%

intense profanity and expressions of frustration

Expletives followed by a noun or pronoun

swear words and expletives

No Comments

No Known Activations