INDEX

Explanations

keywords and phrases related to specific actions and evaluations in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

 հղումներ

-0.65

 nakalista

-0.59

NewUrlParser

-0.59

RTDA

-0.55

цездатний

-0.53

 kaarangay

-0.52

 Administrativna

-0.52

 Italijanski

-0.50

Ɓ

-0.50

MessageTagHelper

-0.49

POSITIVE LOGITS

ReusableCell

0.55

 CreateTagHelper

0.48

SequentialGroup

0.48

routeProvider

0.43

 AppCompatTheme

0.40

ParallelGroup

0.40

openConnection

0.38

Browsable

0.38

 dậy

0.37

Géographie

0.36

Activations Density 0.293%

keywords and phrases related to specific actions and evaluations in various contexts

No Comments

No Known Activations