INDEX

Explanations

references to quantities, positions, or identifiers within contexts

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

fourth directory or picture

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

PerformLayout

-0.67

XmlAccessType

-0.60

libft

-0.59

RectangleBorder

-0.57

elemField

-0.56

SourceChecksum

-0.55

 uLocal

-0.55

oredCriteria

-0.55

hObject

-0.54

AsUp

-0.54

POSITIVE LOGITS

Cyfeiriadau

0.43

Билгалдахарш

0.41

zarchiwizowane

0.39

__(/*!

0.36

 препратки

0.35

AutoScaleMode

0.35

Спољашње

0.34

 vostri

0.34

Personensuche

0.33

 ligiloj

0.33

Activations Density 0.061%

references to quantities, positions, or identifiers within contexts

fourth directory or picture

No Comments

No Known Activations