INDEX

Explanations

structured data or coding syntax

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

list items or attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

ItemBackground

-0.53

Literatuur

-0.46

FieldDescriptor

-0.41

 Waste

-0.40

addWidget

-0.39

TextMessage

-0.38

Whis

-0.38

 acque

-0.37

Waste

-0.36

Toxicity

-0.36

POSITIVE LOGITS

 CreateTagHelper

0.55

rungsseite

0.54

GEBURTSDATUM

0.52

 चीज़ों

0.49

Vidite

0.47

 المعيارى

0.45

Życiorys

0.45

ValueStyle

0.44

 Мексичка

0.43

NameInMap

0.42

Activations Density 0.005%

structured data or coding syntax

list items or attributes

No Comments

No Known Activations