INDEX

Explanations

terms and phrases related to location and changes in positioning

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

ほう

-0.40

Pragma

-0.39

omir

-0.39

 bearers

-0.39

 resi

-0.39

 Lamp

-0.38

 motor

-0.38

➌

-0.37

 full

-0.37

 fullness

-0.37

POSITIVE LOGITS

AutoScaleMode

0.57

########.

0.52

ControllerAdvice

0.51

enumii

0.50

 autorytatywna

0.48

tagext

0.47

fromnode

0.46

Controllo

0.46

katapos

0.44

addWidget

0.43

Activations Density 0.712%

terms and phrases related to location and changes in positioning

No Comments

No Known Activations