INDEX

Explanations

references to reusable cell structures in programming or data handling contexts

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

 rule

-0.46

/**

-0.44

 Sheikh

-0.41

 Linton

-0.40

 Generales

-0.40

 contextLoads

-0.40

 self

-0.40

 noastre

-0.40

 anklicken

-0.40

mtr

-0.39

POSITIVE LOGITS

 BeautifulSoup

3.28

jsoup

0.90

ScopeManager

0.88

Soup

0.84

soup

0.82

 Soup

0.69

TypedDataSet

0.65

 soup

0.65

 Administrativna

0.63

uxxxx

0.62

Activations Density 0.000%

references to reusable cell structures in programming or data handling contexts

No Comments

No Known Activations