INDEX

Explanations

technical terms and programming-related expressions

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

types or errors

np_acts-logits-general · gemini-2.0-flash

type errors and undefined

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 InputDecoration

-0.63

PositiveButton

-0.53

 AttributeSet

-0.52

InitStruct

-0.51

Headline

-0.49

TestingModule

-0.49

BeginInit

-0.48

sement

-0.47

 Headline

-0.46

Réalisation

-0.46

POSITIVE LOGITS

typeof

0.95

 typeof

0.95

 utafitiHapana

0.84

 isinstance

0.79

typeOf

0.71

 Infórmanos

0.64

 isKindOfClass

0.63

 TypeError

0.62

cetype

0.61

 type

0.61

Activations Density 0.297%

technical terms and programming-related expressions

types or errors

type errors and undefined

No Comments

No Known Activations