INDEX

Explanations

elements related to parameters and their attributes in configuration settings

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

"param" followed by "name"

np_acts-logits-general · gemini-2.0-flash

param name attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

__*/

-0.75

SharedCtor

-0.67

 tartalomajánló

-0.67

expandindo

-0.65

AsUp

-0.63

 createContext

-0.62

tagHelper

-0.61

الحياه

-0.60

glEnd

-0.60

cerely

-0.59

POSITIVE LOGITS

param

0.83

{*}

0.66

 name

0.64

gr

0.55

 param

0.54

اعم

0.54

PARAM

0.52

 NAME

0.49

tre

0.48

 victim

0.48

Activations Density 0.029%

elements related to parameters and their attributes in configuration settings

"param" followed by "name"

param name attributes

No Comments

No Known Activations