INDEX

Explanations

mathematical structures and formatting elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 EdgeInsets

-0.58

Wak

-0.58

Datuak

-0.57

 समीक्षक

-0.53

Wak

-0.52

[-\

-0.50

 Мекси

-0.49

 lieu

-0.49

 Wain

-0.49

entas

-0.48

POSITIVE LOGITS

\{\\

0.88

 للاسماء

0.81

mtable

0.71

AddHtmlAttribute

0.67

AccessorTable

0.66

KURZBESCHREIBUNG

0.66

 sélectionné

0.66

soever

0.65

 Перейти

0.65

 pulito

0.65

Activations Density 0.036%

mathematical structures and formatting elements

No Comments

No Known Activations