INDEX

Explanations

references to web-related technologies and frameworks

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

Opening lines of emails or messages

np_acts-logits-general · gemini-2.0-flash

greetings followed by names

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

ayar

-0.53

何より

-0.52

Jvm

-0.50

aarrggbb

-0.50

きましょう

-0.49

CAPTION

-0.45

IANGLES

-0.45

 stand

-0.44

bacillus

-0.44

だからこそ

-0.43

POSITIVE LOGITS

Hi

1.90

Hello

1.83

 Hello

1.81

Hi

1.80

 hello

1.62

Dear

1.55

hello

1.44

 Dear

1.41

Hey

1.40

hi

1.39

Activations Density 0.660%

references to web-related technologies and frameworks

Opening lines of emails or messages

greetings followed by names

No Comments

No Known Activations