INDEX

Explanations

phrases related to accusations, negativity, and conspiracy theories

oai_token-act-pair · gpt-3.5-turbo

content related to conspiracy theories and social manipulation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

etooth

-0.76

oulder

-0.75

ecided

-0.74

oji

-0.73

Enlarge

-0.73

 Towns

-0.71

 bumps

-0.71

ulty

-0.71

 outdoors

-0.69

Ç

-0.69

POSITIVE LOGITS

 Illuminati

1.32

 disinformation

1.29

 treason

1.27

 deception

1.21

 Satanic

1.21

 orchestrated

1.20

 deceit

1.19

 collusion

1.18

 nefarious

1.17

 blackmail

1.15

Activations Density 0.994%

phrases related to accusations, negativity, and conspiracy theories

content related to conspiracy theories and social manipulation

No Comments

No Known Activations