INDEX

Explanations

specific technical terms and identifiers related to system calls and file attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 vueltas

-0.44

 mármol

-0.43

 aikaa

-0.42

 feroit

-0.40

insatz

-0.39

 rodillas

-0.38

 leçon

-0.38

ksesta

-0.38

actéristi

-0.37

 muerto

-0.37

POSITIVE LOGITS

ux

0.66

يكب

0.59

uite

0.59

uxx

0.56

UCE

0.56

uve

0.56

uom

0.55

UCA

0.54

urable

0.53

VersionUID

0.53

Activations Density 2.007%

specific technical terms and identifiers related to system calls and file attributes

No Comments

No Known Activations