INDEX

Explanations

highlights or mentions of specific locations or positions

oai_token-act-pair · gpt-3.5-turbo

references to positions or locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

yss

-0.73

ramid

-0.72

 funeral

-0.69

otype

-0.66

 restraint

-0.66

 backbone

-0.65

IFE

-0.63

ository

-0.63

issance

-0.63

gh

-0.62

POSITIVE LOGITS

 spots

1.36

pots

1.28

lights

1.02

hips

0.99

bugs

0.90

pread

0.90

bucks

0.86

 spot

0.85

hot

0.83

seeing

0.82

Activations Density 0.008%

highlights or mentions of specific locations or positions

references to positions or locations

No Comments

No Known Activations