INDEX

Explanations

elements related to social media content, specifically Twitter posts

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

-0.71

expandindo

-0.65

oredCriteria

-0.56

ükemmel

-0.50

 HasFactory

-0.50

tagHelperRunner

-0.49

TagMode

-0.48

 <=",

-0.48

migrationBuilder

-0.47

 Normdatei

-0.47

POSITIVE LOGITS

 Instagram

0.63

Instagram

0.52

 Facebook

0.48

 instagram

0.46

 INSTAGRAM

0.45

 Twitter

0.44

 Numerade

0.44

Facebook

0.41

 tweeting

0.40

instagram

0.40

Activations Density 0.001%

elements related to social media content, specifically Twitter posts

No Comments

No Known Activations