INDEX

Explanations

references to "up" or "down"

oai_token-act-pair · gemini-2.0-flash

down

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

Not in Any Lists

No Comments

Negative Logits

 down

-1.31

down

-1.24

Down

-1.14

 Down

-1.13

 DOWN

-1.07

DOWN

-1.00

 downward

-0.75

downs

-0.75

 kasarigan

-0.71

up

-0.71

POSITIVE LOGITS

 الحره

0.66

 للمعارف

0.65

 disambiguazione

0.63

ंदीखरीदारी

0.62

\{\\

0.59

Carriera

0.53

UnitTesting

0.53

versiti

0.52

 Eindruck

0.51

❷

0.51

Activations Density 2.966%

references to "up" or "down"

down

No Comments

No Known Activations