INDEX

Explanations

patterns in numerical or coded data structures

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

Dates separated by hyphens

np_acts-logits-general · gemini-2.0-flash

dates and numbers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

期刊论文

-0.51

vœ

-0.50

 disfru

-0.42

 arşivlendi

-0.40

 coupable

-0.39

ambién

-0.39

 verrou

-0.39

 senderos

-0.37

 cocinar

-0.37

erschied

-0.37

POSITIVE LOGITS

 Meksiku

0.75

 Taktlose

0.75

__*/

0.68

AndEndTag

0.65

migrationBuilder

0.64

 transfieras

0.63

libft

0.63

complexContent

0.63

ftagPool

0.62

SequentialGroup

0.61

Activations Density 0.193%

patterns in numerical or coded data structures

Dates separated by hyphens

dates and numbers

No Comments

No Known Activations