INDEX

Explanations

references to the shape and morphology of objects or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wój

-0.92

CodedInputStream

-0.89

 myſelf

-0.89

 Monfieur

-0.87

$_"

-0.87

 Munro

-0.87

 thirds

-0.85

soever

-0.85

 acorns

-0.85

 photolibrary

-0.83

POSITIVE LOGITS

 shape

1.87

 SHAPE

1.81

 shapes

1.73

 Shape

1.71

shape

1.67

SHAPE

1.65

 Shapes

1.64

Shape

1.63

shapes

1.55

 shaped

1.43

Activations Density 0.055%

references to the shape and morphology of objects or entities

No Comments

No Known Activations