INDEX

Explanations

variations of "win" or "won't"

oai_token-act-pair · gemini-2.0-flash

won

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_20/width_16k/average_l0_11

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

IsContent

-0.81

URLException

-0.72

Efq

-0.69

 فريبيس

-0.68

 Савезне

-0.66

eryllium

-0.66

astify

-0.65

 Infórmanos

-0.65

Personensuche

-0.63

+#+#

-0.63

POSITIVE LOGITS

0.60

nt

0.59

0.56

0.55

et

0.54

not

0.52

not

0.52

ent

0.49

0.47

gh

0.47

Activations Density 0.623%

variations of "win" or "won't"

won

No Known Activations