INDEX

Explanations

programming-related terms and structured data elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 pères

-0.45

antaranya

-0.44

 carrés

-0.41

tedì

-0.40

 numéros

-0.40

้อมูล

-0.39

warnai

-0.38

nextProps

-0.38

 réunions

-0.38

šanās

-0.38

POSITIVE LOGITS

PreInfinity

0.66

ScopeManager

0.55

 MonoBehaviour

0.54

/**

0.53

 UnityEngine

0.52

 TMPro

0.51

XmlType

0.50

 iParam

0.50

UnusedPrivate

0.49

/*

0.48

Activations Density 0.116%

programming-related terms and structured data elements

No Comments

No Known Activations