INDEX

Explanations

terms and phrases related to safety and protection in emergency scenarios

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 Paglinawan

-0.46

omitempty

-0.37

ConstraintMaker

-0.37

 nadzieję

-0.37

apeake

-0.36

 jakości

-0.35

wußt

-0.33

ghouse

-0.33

essentiel

-0.32

quality

-0.31

POSITIVE LOGITS

 dangerous

0.76

 unsafe

0.71

 danger

0.71

 المعيارى

0.71

 safety

0.71

 Safety

0.71

 hazardous

0.70

Safety

0.68

 hazard

0.66

 safely

0.66

Activations Density 0.306%

terms and phrases related to safety and protection in emergency scenarios

No Comments

No Known Activations