INDEX

Explanations

mentions of LGBTQ+ identities, particularly lesbian and bisexual individuals

oai_token-act-pair · gpt-3.5-turbo

references to lesbian and bisexual identities within the context of LGBTQ+ issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

frames

-0.90

alez

-0.89

æĸ¹

-0.82

enegger

-0.81

ULE

-0.81

osuke

-0.78

alty

-0.78

anim

-0.75

uden

-0.74

lamm

-0.74

POSITIVE LOGITS

 couples

0.96

ism

0.93

 separat

0.83

sex

0.83

 bisexual

0.78

 porn

0.78

 sisters

0.77

 romance

0.73

 lesbian

0.72

 nightclub

0.72

Activations Density 0.024%

mentions of LGBTQ+ identities, particularly lesbian and bisexual individuals

references to lesbian and bisexual identities within the context of LGBTQ+ issues

No Comments

No Known Activations