INDEX

Explanations

words related to personal preferences or rankings, and words incidentally appearing near them

oai_token-act-pair · gemini-2.0-flash

preferences

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_25/width_16k/average_l0_41

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

AndEndTag

-0.77

VersionUID

-0.77

twimg

-0.71

########.

-0.67

TargetException

-0.65

printStackTrace

-0.63

ніципалі

-0.62

 AssemblyCulture

-0.61

DebuggerNonUser

-0.60

 favorite

-0.59

POSITIVE LOGITS

 boys

0.53

 Norvège

0.49

men

0.48

président

0.46

zeln

0.46

 children

0.45

 Nose

0.44

 seas

0.44

 bocas

0.44

 filles

0.43

Activations Density 6.899%

words related to personal preferences or rankings, and words incidentally appearing near them

preferences

No Comments

No Known Activations