INDEX

Explanations

words related to profanity or actions that are considered vulgar or offensive

oai_token-act-pair · gpt-3.5-turbo

terms related to professional conduct or professionalism

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

warm

-0.88

forth

-0.76

 lodge

-0.71

warming

-0.71

MENTS

-0.70

wolves

-0.67

nuts

-0.67

-0.65

witz

-0.64

LESS

-0.63

POSITIVE LOGITS

iled

1.17

aned

1.10

iler

1.09

 prof

1.03

ilers

1.02

iles

1.01

iling

1.01

essor

0.99

usion

0.96

anity

0.96

Activations Density 0.017%

words related to profanity or actions that are considered vulgar or offensive

terms related to professional conduct or professionalism

No Comments

No Known Activations