INDEX

Explanations

Japanese company names

oai_token-act-pair · gemini-2.0-flash

terms related to categories or designations

np_token-act-pair-logits · gpt-4o-mini

Japanese descriptions

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Majefty

-0.86

 purpoſe

-0.84

 pleaſure

-0.84

 houſe

-0.84

 Houſe

-0.79

 paſſ

-0.79

 Monfieur

-0.78

 ſever

-0.77

 uſe

-0.76

 laſt

-0.75

POSITIVE LOGITS

SuccessListener

0.52

 للاسماء

0.51

 محفوظة

0.47

hen

0.47

PreferredItem

0.44

echo

0.43

enumi

0.42

soort

0.42

Res

0.42

 مرئيه

0.42

Activations Density 0.311%

Japanese company names

terms related to categories or designations

Japanese descriptions

No Known Activations