INDEX

Explanations

occurrences of numerical data and symbols related to power and failures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Datuak

-0.78

enumi

-0.73

 &___

-0.69

ThemeOverlay

-0.69

:✨

-0.68

AndEndTag

-0.66

 tartalomajánló

-0.64

rungsseite

-0.63

niająca

-0.62

 AssemblyTitle

-0.62

POSITIVE LOGITS

0.79

 sixteenth

0.68

 sixteen

0.67

 Sixteen

0.63

 Sixteenth

0.62

XVI

0.61

XVI

0.57

۱۶

0.54

 sexta

0.54

Activations Density 0.231%

occurrences of numerical data and symbols related to power and failures

No Comments

No Known Activations