INDEX

Explanations

phrases indicating proximity or closeness

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

 AssemblyCulture

-0.45

 sorella

-0.41

 typelib

-0.38

WaitGroup

-0.37

 letteratura

-0.35

 composizione

-0.35

 myö

-0.35

 Imperio

-0.35

 Talley

-0.34

 generazione

-0.34

POSITIVE LOGITS

Near

0.94

near

0.92

 Near

0.89

 near

0.81

NEAR

0.80

 NEAR

0.77

 صوتيه

0.75

接近

0.70

OrEqual

0.68

 Nearest

0.65

Activations Density 0.021%

phrases indicating proximity or closeness

No Comments

No Known Activations