INDEX

Explanations

language related to expressing a specific point of view or belief

oai_token-act-pair · gpt-3.5-turbo

phrases related to opinions and positions on various issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

NetMessage

-1.15

issance

-0.73

STON

-0.73

 Sabha

-0.69

duct

-0.66

 Explos

-0.65

amaz

-0.64

ãĥ¼ãĥĨãĤ£

-0.63

Luck

-0.63

nice

-0.63

POSITIVE LOGITS

 stances

1.27

 stance

1.27

 positions

0.98

holders

0.94

 position

0.91

 views

0.90

 beliefs

0.89

 viewpoint

0.88

 regarding

0.87

 opinions

0.87

Activations Density 0.075%

language related to expressing a specific point of view or belief

phrases related to opinions and positions on various issues

No Comments

No Known Activations

language related to expressing a specific point of view or belief

phrases related to opinions and positions on various issues

No Comments

No Known Activations