INDEX

Explanations

characters and formatting related to programming code and syntax

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 получается

-0.39

ARROLL

-0.37

 appelez

-0.34

 Formen

-0.33

 aquello

-0.32

occasione

-0.32

 siku

-0.32

Hozzáférés

-0.32

 αποτε

-0.30

 similarly

-0.30

POSITIVE LOGITS

 المعيارى

0.86

OGND

0.73

 AssemblyCompany

0.73

 يتيمه

0.69

__(/*!

0.61

DeleteBehavior

0.57

$_(

0.57

ंदीखरीदारी

0.56

 témoig

0.54

 kasarigan

0.54

Activations Density 0.003%

characters and formatting related to programming code and syntax

No Comments

No Known Activations