INDEX

Explanations

references to growth, potential, and improvement in various contexts

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

measurement after time

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

Advertisement

-0.40

 during

-0.39

 hanger

-0.38

]-->

-0.37

 während

-0.37

newtheorem

-0.37

 durante

-0.36

 lors

-0.35

 lär

-0.35

WireFormatLite

-0.35

POSITIVE LOGITS

OGND

0.57

WebVitals

0.54

 AssemblyCulture

0.47

eleste

0.45

featureID

0.44

RTLR

0.42

OptionsMenu

0.41

pushFollow

0.41

っこう

0.41

istani

0.41

Activations Density 0.049%

references to growth, potential, and improvement in various contexts

measurement after time

No Comments

No Known Activations