INDEX

Explanations

punctuation marks and formatting elements in written text

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

code comments and list items

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

 lắm

-0.38

]}>

-0.38

 zufolge

-0.37

 Ergän

-0.35

_]

-0.35

Tom

-0.35

 opcional

-0.35

 Ferreira

-0.35

extras

-0.35

 eterno

-0.34

POSITIVE LOGITS

ValueStyle

0.69

 resourceCulture

0.68

DoubleQuotes

0.65

 للاسماء

0.64

httphttps

0.63

 للمعارف

0.59

 HasFactory

0.57

LookAnd

0.56

rungsseite

0.56

NameInMap

0.56

Activations Density 0.059%

punctuation marks and formatting elements in written text

code comments and list items

No Comments

No Known Activations