INDEX

Explanations

phrases and language related to clarity and explicitness in communication

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

explicitly state or indicate information

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

 surla

-0.56

انجليز

-0.42

Dichloropropane

-0.42

getLog

-0.41

'{@

-0.40

getLogger

-0.40

createClass

-0.37

ustice

-0.37

 ContentValues

-0.36

 unknownFields

-0.36

POSITIVE LOGITS

 explicitly

0.64

 expressly

0.56

 explí

0.52

 specifically

0.50

 sumpay

0.50

 intptr

0.48

 ausdrücklich

0.47

explicit

0.47

 note

0.46

 clairement

0.46

Activations Density 0.051%

phrases and language related to clarity and explicitness in communication

explicitly state or indicate information

No Comments

No Known Activations