INDEX

Explanations

keywords related to attitudes and perceptions

oai_token-act-pair · gpt-3.5-turbo

references to societal attitudes and perceptions

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

avez

-0.78

addafi

-0.75

ded

-0.73

Delivery

-0.70

enary

-0.67

amen

-0.66

ÄŁ

-0.66

Jub

-0.66

aman

-0.65

gran

-0.64

POSITIVE LOGITS

 attitudes

1.02

pring

0.86

ocial

0.85

hips

0.82

insula

0.82

hip

0.80

 guiActiveUn

0.79

uggest

0.77

 toward

0.73

mith

0.72

Activations Density 0.022%

keywords related to attitudes and perceptions

references to societal attitudes and perceptions

No Comments

No Known Activations