INDEX

Explanations

phrases related to expressing a stance or viewpoint

oai_token-act-pair · gpt-3.5-turbo

instances of the word "position" and its context in discussions of stance or viewpoint

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

-0.87

issance

-0.78

rosso

-0.75

ãĥ¼ãĥĨãĤ£

-0.74

ebook

-0.74

displayText

-0.73

STON

-0.72

anton

-0.69

cia

-0.68

 Explos

-0.67

POSITIVE LOGITS

holders

1.16

 positions

1.16

 position

1.14

holder

1.03

 stances

0.91

 holders

0.88

 stance

0.86

quo

0.84

 holder

0.80

senal

0.76

Activations Density 0.028%

phrases related to expressing a stance or viewpoint

instances of the word "position" and its context in discussions of stance or viewpoint

No Comments

No Known Activations