INDEX

Explanations

phrases related to contrasting perspectives or ideas

oai_token-act-pair · gpt-3.5-turbo

complex interactions and relationships between concepts and events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

osa

-0.63

Zah

-0.59

Zac

-0.58

>>>

-0.58

tower

-0.56

hart

-0.55

meet

-0.55

 Originally

-0.54

Ult

-0.54

Hud

-0.54

POSITIVE LOGITS

 nonetheless

1.16

etheless

0.83

 nevertheless

0.79

 awfully

0.76

 darn

0.74

 damn

0.73

 oneself

0.72

 impro

0.66

 downright

0.66

ufact

0.66

Activations Density 0.738%

phrases related to contrasting perspectives or ideas

complex interactions and relationships between concepts and events

No Comments

No Known Activations