INDEX

Explanations

negation or expressions of disapproval

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

opening parenthesis

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

yntaxException

-0.44

exitRule

-0.44

 Marín

-0.42

 itſelf

-0.42

marathon

-0.42

 betweenstory

-0.40

 Bingham

-0.40

LookAnd

-0.40

mphony

-0.40

 asynchronously

-0.38

POSITIVE LOGITS

#+#

0.48

IsContent

0.46

 بيها

0.42

(!__

0.42

 chance

0.41

쯤

0.41

setopt

0.40

ButtonModule

0.40

 дописавши

0.39

NameInMap

0.39

Activations Density 0.009%

negation or expressions of disapproval

opening parenthesis

No Comments

No Known Activations