INDEX

Explanations

words that suggest a quick or partial view or insight into something

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate hints or glimpses of what is to come

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

die

-0.82

ials

-0.75

udeb

-0.74

glomer

-0.73

share

-0.71

atography

-0.70

FactoryReloaded

-0.70

scill

-0.68

ansom

-0.66

Mub

-0.66

POSITIVE LOGITS

 glimps

0.94

 riches

0.85

 greatness

0.78

 doom

0.78

 heights

0.78

 hints

0.78

 beginnings

0.77

 brilliance

0.76

 whats

0.75

 impending

0.73

Activations Density 0.190%

words that suggest a quick or partial view or insight into something

phrases that indicate hints or glimpses of what is to come

No Comments

No Known Activations