INDEX

Explanations

words related to the LGBTQ+ community, especially terms related to homosexuality

oai_token-act-pair · gpt-3.5-turbo

references to homosexuality and related topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

UGE

-0.77

 shroud

-0.71

ãģį

-0.68

hower

-0.67

FUL

-0.66

 canopy

-0.66

 sender

-0.63

 indirect

-0.61

FORE

-0.61

AMS

-0.61

POSITIVE LOGITS

estead

1.43

emade

1.36

ework

1.19

osexual

1.12

eless

1.09

ogeneous

1.04

icide

1.04

ogenous

1.03

neys

1.02

ocaust

1.01

Activations Density 0.009%

words related to the LGBTQ+ community, especially terms related to homosexuality

references to homosexuality and related topics

No Comments

No Known Activations