INDEX

Explanations

patterns related to programming or scripting code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Wikimedijinoj

-0.82

__":

-0.78

+#+#

-0.76

 queſta

-0.76

__':

-0.76

ChildScrollView

-0.76

ArrowToggle

-0.75



-0.74

ſſung

-0.73

IsMutable

-0.72

POSITIVE LOGITS

//

0.35

//

0.34

Mr

0.33

<strong>

0.33

↵↵

0.32

if

0.32

↵

0.31

↵↵↵

0.30

return

0.30

<eos>

0.28

Activations Density 0.017%

patterns related to programming or scripting code

No Comments

No Known Activations