INDEX

Explanations

names or mentions of individuals

oai_token-act-pair · gpt-3.5-turbo

the pronoun "I"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Valkyrie

-0.73

*/(

-0.72

 convol

-0.72

 Haunted

-0.71

skirts

-0.71

 alternate

-0.70

 colleg

-0.68

 Compass

-0.67

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.67

 mileage

-0.66

POSITIVE LOGITS

ye

1.16

orno

1.12

ogi

1.05

acs

1.05

edi

1.03

oga

0.99

uli

0.99

anni

0.99

imi

0.98

wa

0.97

Activations Density 0.045%

names or mentions of individuals

the pronoun "I"

No Comments

No Known Activations