INDEX

Explanations

HTML metadata tags and attributes

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

meta tags

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

олові

-0.76

parsedMessage

-0.75

featureID

-0.68

AndroidJUnit

-0.66

tagext

-0.65

:✨

-0.61

fjspx

-0.60

enterOuterAlt

-0.59

 onPostExecute

-0.59

msgTypes

-0.58

POSITIVE LOGITS

meta

0.68

 meta

0.52

Meta

0.52

メタ

0.47

metas

0.47

 méta

0.46

 Meta

0.45

META

0.44

 metaphysical

0.40

 metas

0.40

Activations Density 0.002%

HTML metadata tags and attributes

meta tags

No Comments

No Known Activations