INDEX

Explanations

phrases related to the concept of correctness, specifically political correctness

oai_token-act-pair · gpt-3.5-turbo

references to political correctness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Organ

-0.77

 Rider

-0.70

joining

-0.67

 Wings

-0.67

ogo

-0.66

paying

-0.65

 Scher

-0.64

 Revenue

-0.64

 participating

-0.63

Bus

-0.62

POSITIVE LOGITS

 correctness

1.38

 orthodoxy

0.99

 guiActiveUn

0.95

 behavi

0.91

 prejudice

0.86

geist

0.85

 ignorance

0.85

icity

0.84

iveness

0.81

iquette

0.80

Activations Density 0.014%

phrases related to the concept of correctness, specifically political correctness

references to political correctness

No Comments

No Known Activations