INDEX

Explanations

formatting or structural elements in the text

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

Non-English words and symbols

np_acts-logits-general · gemini-2.0-flash

email metadata

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

the

-0.65

-0.53

uny

-0.52

 some

-0.52

 several

-0.51

".

-0.50

citet

-0.49

 various

-0.48

 either

-0.48

few

-0.47

POSITIVE LOGITS

 мәкал

1.00

complexContent

0.98

0.90

uxxxx

0.89

 disambiguazione

0.86

 Chwiliwch

0.83

 onCancelled

0.82

 Meksiku

0.79

InjectAttribute

0.79

Geplaatst

0.78

Activations Density 0.414%

formatting or structural elements in the text

Non-English words and symbols

email metadata

No Comments

No Known Activations