INDEX

Explanations

mentions of negative social behaviors and interactions

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

postmedia

-0.46

 desembar

-0.41

NUMX

-0.41

RTDA

-0.41

dư

-0.40

GEBURTSDATUM

-0.40

 Paddle

-0.39

髦

-0.39

 Autor

-0.39

Premise

-0.39

POSITIVE LOGITS

 competitors

0.46

Competi

0.41

 ModelExpression

0.41

 competitor

0.40

Competition

0.40

prometheus

0.39

 competing

0.39

üyada

0.38

 конкурен

0.38

Competitive

0.37

Activations Density 0.017%

mentions of negative social behaviors and interactions

No Comments

No Known Activations