INDEX

Explanations

concepts related to purity and cleansing

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

pure, purity, purifying

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

MENAFN

-0.67

awtextra

-0.65

RegistryLite

-0.61

MLLoader

-0.58

 utafitiHapana

-0.54

 despatch

-0.52

UpInside

-0.51

 Widodo

-0.49

Crossref

-0.48

 ModelExpression

-0.47

POSITIVE LOGITS

 purity

1.02

 pureza

0.89

 Purity

0.84

Purity

0.78

 purer

0.73

 purification

0.73

 purify

0.70

 clean

0.69

 pure

0.68

 Purification

0.66

Activations Density 0.020%

concepts related to purity and cleansing

pure, purity, purifying

No Comments

No Known Activations