INDEX

Explanations

elements related to numerical or programming constructs, possibly indicating code or data structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

 tartalomajánló

-0.70

 мәкал

-0.63

SequentialGroup

-0.61

 Chwiliwch

-0.58

 ویکی‌پدی

-0.56

awtextra

-0.54

$_"

-0.54

 OFDb

-0.52

 NSCoder

-0.51

WarningLevel

-0.51

POSITIVE LOGITS

Források

0.36

 helse

0.33

 odkazy

0.33

 samarbeid

0.32

 krep

0.31

taan

0.31

RESUMO

0.30

 Socialista

0.29

 république

0.29

 inyección

0.29

Activations Density 1.052%

elements related to numerical or programming constructs, possibly indicating code or data structures

No Comments

No Known Activations