INDEX

Explanations

references to subsections and numbered lists within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

-0.71

var

-0.67

//

-0.62

gu

-0.61

var

-0.60

-0.59

 الع

-0.58

Gu

-0.57

сс

-0.57

POSITIVE LOGITS

subsection

2.51

subsubsection

2.04

 himſelf

1.42

 itſelf

1.40

 myſelf

1.40

BibitemShut

1.26

 ་་

1.24

 ―――――

1.23

 autorytatywna

1.20

 Jefus

1.20

Activations Density 0.033%

references to subsections and numbered lists within the text

No Comments

No Known Activations