INDEX

Explanations

specific phrases indicating conditions or legal contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 noastre

-0.36

 nostru

-0.35

 meilleur

-0.31

zusetzen

-0.31

ėjimo

-0.28

 розы

-0.28

████

-0.28

 песню

-0.28

zoic

-0.27

 działal

-0.27

POSITIVE LOGITS

nodoc

0.57

SourceChecksum

0.57

Vidite

0.54

parsedMessage

0.54

хьтан

0.52

 kaarangay

0.52

RectangleBorder

0.51

 CURIAM

0.51

utilisons

0.50

EndContext

0.50

Activations Density 1.233%

specific phrases indicating conditions or legal contexts

No Comments

No Known Activations