INDEX

Explanations

mentions of physical or metaphorical locations or positions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating placement or position

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

 trick

-0.69

++)

-0.66

gery

-0.65

eed

-0.63

++

-0.61

 loss

-0.60

lig

-0.59

 shampoo

-0.59

isin

-0.59

naires

-0.59

POSITIVE LOGITS

 pedest

1.03

 hierarchy

0.77

Interstitial

0.77

onomous

0.73

OPLE

0.72

 prestigious

0.70

 hierarchical

0.69

thumbnails

0.68

 totem

0.67

 jeopardy

0.67

Activations Density 0.374%

mentions of physical or metaphorical locations or positions

phrases indicating placement or position

No Comments

No Known Activations