INDEX

Explanations

references to options or choices presented in a document

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

the

-0.49

The

-0.49

rmtree

-0.46

‘

-0.43

</tr>

-0.42

Her

-0.42

Hud

-0.40

Be

-0.39

-0.38

POSITIVE LOGITS

 Option

1.32

 OPTION

1.30

Option

1.27

 Options

1.25

 option

1.20

option

1.20

 options

1.16

 OPTIONS

1.16

getOption

1.06

Options

1.05

Activations Density 0.075%

references to options or choices presented in a document

No Comments

No Known Activations