INDEX

Explanations

design-related terms and concepts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "design."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Lauder

-0.81

ieri

-0.80

 Sheen

-0.80

ICAN

-0.78

nikov

-0.78

selves

-0.75

Ago

-0.74

 Arist

-0.72

 Adin

-0.71

lehem

-0.71

POSITIVE LOGITS

ators

1.09

ator

1.03

yout

1.01

ations

0.94

ating

0.89

 design

0.88

 flaw

0.86

ated

0.85

 designs

0.84

ates

0.83

Activations Density 0.040%

design-related terms and concepts

instances of the word "design."

No Comments

No Known Activations