INDEX

Explanations

references to LGBTQ+ topics

oai_token-act-pair · gpt-3.5-turbo

terms related to gay identity and issues

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

urers

-0.76

è¦ļéĨĴ

-0.71

âĵĺ

-0.68

arily

-0.67

Condition

-0.67

Reviewer

-0.66

Dur

-0.65

)=(

-0.64

PsyNetMessage

-0.62

effective

-0.61

POSITIVE LOGITS

 marriage

1.00

dar

0.97

atri

0.95

lord

0.93

glers

0.93

 pride

0.91

 porn

0.90

 couples

0.88

 rights

0.88

 slurs

0.87

Activations Density 0.024%

references to LGBTQ+ topics

terms related to gay identity and issues

No Comments

No Known Activations