INDEX

Explanations

the presence of tags or indicators at the beginning of a text segment

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 TextAppearance

-0.60

 disambiguazione

-0.59

 ISNI

-0.58

 nahilalakip

-0.58

Diwedd

-0.56

lioz

-0.52

 kasarigan

-0.48

 ProtoMessage

-0.48

 forth

-0.48

SourceChecksum

-0.47

POSITIVE LOGITS

BeginContext

0.67

 Unmarshal

0.55

Portail

0.53

 Amen

0.52

例句

0.52

 Methuen

0.52

RTRS

0.52

}$\

0.51

Amen

0.51

uxxxx

0.50

Activations Density 0.011%

the presence of tags or indicators at the beginning of a text segment

No Comments

No Known Activations