INDEX

Explanations

references to visuals or imagery

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

각주

-0.52

PreInfinity

-0.47

 Jones

-0.44

enzyme

-0.43

nsw

-0.42

 JONES

-0.42

 Neves

-0.40

farm

-0.40

ội

-0.40

 okuyayım

-0.39

POSITIVE LOGITS

 picture

2.05

picture

1.80

 Picture

1.72

Picture

1.63

 PICTURE

1.53

 pictures

1.45

PICTURE

1.45

pictures

1.25

Pictures

1.25

 Pictures

1.25

Activations Density 0.068%

references to visuals or imagery

No Comments

No Known Activations