INDEX

Explanations

concepts related to the interaction and connection between words and their meanings

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

how I

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

RenderAtEndOf

-0.80

UserScript

-0.72

transQ

-0.72

tonode

-0.69

ViewImports

-0.68

StoryboardSegue

-0.67

featureID

-0.65

SequentialGroup

-0.65

 kasarigan

-0.64

 المعيارى

-0.64

POSITIVE LOGITS

 subtle

0.42

 material

0.38

 sutil

0.38

细节

0.36

 paysages

0.35

 texturas

0.34

 subtly

0.33

 Herramientas

0.33

的关系

0.32

 materials

0.32

Activations Density 0.019%

concepts related to the interaction and connection between words and their meanings

how I

No Comments

No Known Activations