INDEX

Explanations

terms associated with radical changes or transformations

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 Verordnung

-0.36

 energética

-0.34

 puntata

-0.32

fka

-0.32

 osserv

-0.31

zab

-0.31

LOGIC

-0.31

 umgeben

-0.31

 acoper

-0.31

 היתה

-0.31

POSITIVE LOGITS

ware

0.89

WARE

0.86

ImageContext

0.79

Ware

0.75

 ware

0.71

 Ware

0.68

 radical

0.68

 Portugal

0.65

scriptId

0.65

printStackTrace

0.64

Activations Density 0.538%

terms associated with radical changes or transformations

No Comments

No Known Activations

terms associated with radical changes or transformations

No Comments

No Known Activations