INDEX

Explanations

phrases related to malicious intent or negative actions of individuals within society

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inspec

-0.46

owulf

-0.46

yntaxException

-0.44

 Infór

-0.43

enumi

-0.43

RTEX

-0.42

 فريبيس

-0.42

urut

-0.42

richTextPanel

-0.40

tagHelperRunner

-0.40

POSITIVE LOGITS

UnusedPrivate

0.55

 PUBLICATIONS

0.42

TypedDataSet

0.39

endpush

0.37

 cherchés

0.37

EXTERN

0.36

LookAnd

0.36

Портали

0.36

PerformLayout

0.35

 abusive

0.35

Activations Density 0.300%

phrases related to malicious intent or negative actions of individuals within society

No Comments

No Known Activations