INDEX

Explanations

initiation tokens or markers, likely signifying the start of a new section or input in 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 @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

Jereo

-0.74

 فريبيس

-0.74

 Administrativna

-0.73

 CreateTagHelper

-0.66

 EconPapers

-0.66

 gynhyrchwyd

-0.64

apimachinery

-0.64

cyklopedia

-0.64

NewUrlParser

-0.63

 BoxFit

-0.63

POSITIVE LOGITS

 quæ

0.49

expandindo

0.49

 habet

0.49

 inſ

0.47

 Effectively

0.47

EndInit

0.46

<unused63>

0.46

 ſtate

0.45

 relazioni

0.45

 blik

0.45

Activations Density 0.262%

initiation tokens or markers, likely signifying the start of a new section or input in the text

No Comments

No Known Activations