INDEX

Explanations

configuration options in code

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

opts and BOOL

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-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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits



-0.78

twimg

-0.64

bootstrapcdn

-0.63

mmation

-0.61

 المعيارى

-0.61

fromnode

-0.61

gonic

-0.60

Bibliograf

-0.59

 Segmentation

-0.58

جمعیت

-0.57

POSITIVE LOGITS

可

0.51

ks

0.50

opts

0.46

ky

0.45

lec

0.41

Opts

0.41

**/

0.40

<?=

0.39

COA

0.39

 opts

0.38

Activations Density 0.022%

configuration options in code

opts and BOOL

No Comments

No Known Activations