INDEX

Explanations

programming language constructs related to class definitions and method declarations

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

 betweenstory

-0.85

 ſind

-0.70

𐄁

-0.69

 queſta

-0.69

parsedMessage

-0.68

 tartalomajánló

-0.68

ésultats

-0.67

 BoxFit

-0.67

<unused3>

-0.66

<unused14>

-0.66

POSITIVE LOGITS

 avanzado

0.29

 comigo

0.28

ter

0.28

verwijspagina

0.27

 vainqueur

0.26

 sentarse

0.25

 conmigo

0.25

 abordagem

0.25

 Anfitrión

0.25

 senhora

0.24

Activations Density 0.003%

programming language constructs related to class definitions and method declarations

No Comments

No Known Activations