INDEX

Explanations

terms and phrases that begin with "pre-" or "post-" and sometimes contain "non-"

oai_token-act-pair · gemini-2.0-flash

pre or post

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 فريبيس

-0.72

новништво

-0.62

 كومونز

-0.58

 CURIAM

-0.53

__((

-0.51

CloseOperation

-0.51

tableFuture

-0.51

putra

-0.50

OREM

-0.50

ynchronously

-0.49

POSITIVE LOGITS

0.64

+#+#

0.61

gan

0.58

GAN

0.55

osis

0.53

LabelTagHelper

0.53

ulation

0.50

EqualTo

0.50

uban

0.49

nan

0.48

Activations Density 1.414%

terms and phrases that begin with "pre-" or "post-" and sometimes contain "non-"

pre or post

No Known Activations