INDEX

Explanations

repeated phrases or the word "same" in different contexts

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

MBOL

-0.37

 objective

-0.37

anair

-0.36

AISSEE

-0.35

iddhar

-0.35

Abbott

-0.35

Cana

-0.34

 kwanza

-0.34

 closer

-0.33

 économiques

-0.33

POSITIVE LOGITS

 derselbe

0.76

 same

0.70

same

0.69

 dieselben

0.68

 derselben

0.63

 dieselbe

0.62

 desselben

0.61

 Paglinawan

0.59

 denselben

0.59

 demselben

0.59

Activations Density 0.028%

repeated phrases or the word "same" in different contexts

No Comments

No Known Activations