INDEX

Explanations

technical terminology and specific scientific or legal terms

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

pecabe

-0.83

transQ

-0.82

 المعيارى

-0.72

aarrggbb

-0.69

 ویکی‌آمباردا

-0.68

InjectAttribute

-0.67

 Administrativna

-0.66

 '\\;'

-0.65

GEBUR

-0.63

 disambiguazione

-0.62

POSITIVE LOGITS

0.43

0.42

<b>

0.40

<strong>

0.40

“

0.39

Activations Density 0.133%

technical terminology and specific scientific or legal terms

No Comments

No Known Activations