INDEX

Explanations

positive affirmations or phrases related to functionality and effectiveness

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

"well as"

np_acts-logits-general · gemini-2.0-flash

as well as

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

lenovo

-0.63

 calendriers

-0.62

">—

-0.57

MENAFN

-0.56

 Hopf

-0.53

ndham

-0.53

 ſtand

-0.52

これも

-0.52

ennen

-0.51

chom

-0.51

POSITIVE LOGITS

as

0.95

CppMethod

0.72

 مرئيه

0.70

AsUp

0.59

ंदीखरीदारी

0.57

=$?

0.57

 као

0.57

})()

0.56

FunctionFlags

0.56

UniformLocation

0.55

Activations Density 0.058%

positive affirmations or phrases related to functionality and effectiveness

"well as"

as well as

No Comments

No Known Activations