INDEX

Explanations

code structure and programming constructs

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

Code-related tokens (const, return, function)

np_acts-logits-general · gemini-2.0-flash

programming and math contexts

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

__':

-0.82

forKey

-0.67

__":

-0.67

XmlAccessorType

-0.67

 समीक्षक

-0.64

WriteTagHelper

-0.62

 Partagez

-0.61

 BoxFit

-0.60

sizeCache

-0.60

matchCondition

-0.59

POSITIVE LOGITS

const

0.89

 const

0.73

useEffect

0.66

return

0.60

 consist

0.50

let

0.50

ro

0.48

ground

0.48

function

0.47

console

0.46

Activations Density 0.023%

code structure and programming constructs

Code-related tokens (const, return, function)

programming and math contexts

No Comments

No Known Activations