INDEX

Explanations

words related to financial corruption or misconduct

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

WebElementEntity

-0.49

 kasarigan

-0.49

toHexString

-0.48

afficheront

-0.47

 choco

-0.43

 autorytatywna

-0.43

setPreferredSize

-0.42

DEPORTES

-0.41

choco

-0.41

transQ

-0.41

POSITIVE LOGITS

ov

0.69

MenuItem

0.57

ov

0.55

 Bean

0.54

OV

0.52

 warrior

0.51

 Marker

0.50

 marker

0.50

Bean

0.50

creek

0.48

Activations Density 0.368%

words related to financial corruption or misconduct

No Comments

No Known Activations