INDEX

Explanations

references to information or situations being revealed or brought to attention

oai_token-act-pair · gpt-3.5-turbo

phrases that discuss bringing issues to attention or highlighting matters of significance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ettings

-0.83

isner

-0.70

CVE

-0.69

 Merrill

-0.67

JUST

-0.67

ESE

-0.66

Dur

-0.65

okia

-0.64

ippi

-0.64

halla

-0.64

POSITIVE LOGITS

 bulb

1.17

weights

1.09

enment

1.07

lights

1.07

 bulbs

1.06

hearted

1.06

ening

1.01

bul

0.97

heartedly

0.96

ened

0.95

Activations Density 0.019%

references to information or situations being revealed or brought to attention

phrases that discuss bringing issues to attention or highlighting matters of significance

No Comments

No Known Activations