INDEX

Explanations

phrases related to vulnerability and susceptibility in various contexts

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

rungsseite

-0.64

 linkovi

-0.56

 MainAxisSize

-0.46

 sprend

-0.40

epa

-0.40

Pozri

-0.40

 saites

-0.40

 supprim

-0.39

 ISNI

-0.38

 poveznice

-0.37

POSITIVE LOGITS

 vulnerable

0.75

 susceptible

0.73

 vulnerability

0.73

 exposed

0.69

 susceptibility

0.66

exposed

0.64

 معرض

0.63

 suscep

0.63

 vulné

0.62

 Vulnerable

0.61

Activations Density 0.128%

phrases related to vulnerability and susceptibility in various contexts

No Comments

No Known Activations