INDEX

Explanations

descriptions of character traits and physical appearances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 poichè

-0.79

 perchè

-0.74

-0.73

(!)

-0.62

pò

-0.61

(!)

-0.59

 Whilst

-0.59

([

-0.59

Whilst

-0.56

(&

-0.56

POSITIVE LOGITS

 twenty

0.69

 seventy

0.68

<eos>

0.67

 fifty

0.66

 fucking

0.66

 thirty

0.65

 forty

0.64

 eighty

0.64

 sixty

0.63

 eighteen

0.63

Activations Density 0.657%

descriptions of character traits and physical appearances

No Comments

No Known Activations