INDEX

Explanations

references to programming and network-related terms or elements

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

 يتيمه

-0.79

 برانيه

-0.59

 >=",

-0.58

 AssemblyTitle

-0.56

antMatchers

-0.55

 ModelRenderer

-0.52

 насељу

-0.52

writeFieldEnd

-0.52

 forState

-0.52

__':

-0.52

POSITIVE LOGITS

 onOptions

0.43

 onCreateView

0.37

 verk

0.35

 szá

0.33

 intellectuelle

0.33

imageio

0.33

 samarbe

0.32

 OnInit

0.31

 return

0.31

PR

0.31

Activations Density 0.106%

references to programming and network-related terms or elements

No Comments

No Known Activations