INDEX

Explanations

phrases referring to specific instances or examples

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

hup

-0.44

LayoutStyle

-0.41

opho

-0.41

atal

-0.41

fjspx

-0.41

ilich

-0.40

留

-0.40

utel

-0.40

pfel

-0.40

 drank

-0.39

POSITIVE LOGITS

存于互联网档案馆

0.57

 tästä

0.57

 this

0.56

 ComVisible

0.50

 этом

0.48

 THIS

0.48

#+#

0.48

Autoritní

0.48

 unknownFields

0.47

Ссылки

0.47

Activations Density 0.263%

phrases referring to specific instances or examples

No Comments

No Known Activations