INDEX

Explanations

characters and numerical values commonly associated with coded or formatted data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 13-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_13/width_16k/average_l0_132

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 information

-0.33

 sudadera

-0.32

-0.31

 lære

-0.31

 arbejde

-0.30

 work

-0.30

 représent

-0.30

 besk

-0.30

 leadership

-0.30

 nation

-0.30

POSITIVE LOGITS

ScopeManager

0.82

 TextInputType

0.78

SequentialGroup

0.77

AndEndTag

0.76

 незавершена

0.75

 MainAxisSize

0.74

parsedMessage

0.71

RectangleBorder

0.70

 mergeFrom

0.70

 transfieras

0.68

Activations Density 0.061%

characters and numerical values commonly associated with coded or formatted data

No Comments

No Known Activations