INDEX

Explanations

countries and entities involved in diplomatic relationships or agreements

oai_token-act-pair · gpt-3.5-turbo

proper nouns and significant entities in a context

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Features by Cosine Similarity

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

Plots

Explanation

Show Test Field

Default Test Text

IFrame

Link

•L9 resid features following 'and'

Negative Logits

tains

-0.79

Angelo

-0.62

mel

-0.61

:,

-0.60

SourceFile

-0.59

 Stam

-0.58

 Â±

-0.58

tes

-0.56

âĶĢâĶĢâĶĢâĶĢ

-0.56

POSITIVE LOGITS

 alike

1.67

are

1.31

 respectively

1.29

 aren

1.20

 were

1.18

 weren

1.12

 mutually

1.04

 collide

1.04

 jointly

1.04

 have

1.02

Activations Density 0.364%

countries and entities involved in diplomatic relationships or agreements

proper nouns and significant entities in a context

No Activations Uploaded