INDEX

Explanations

terms associated with research, disclosure, and scientific information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

-0.78

in

-0.67

 information

-0.67

im

-0.66

 experience

-0.64

an

-0.63

is

-0.61

use

-0.61

to

-0.60

POSITIVE LOGITS

 myſelf

1.59

 ſeveral

1.39

 Monfieur

1.38

 Reſ

1.37

 Majefty

1.34

Eſ

1.33

 feroit

1.32

 whoſe

1.32

 houſe

1.32

 Jefus

1.30

Activations Density 0.323%

terms associated with research, disclosure, and scientific information

No Comments

No Known Activations