INDEX

Explanations

specific days of the week

oai_token-act-pair · gpt-3.5-turbo

mentions of the day "Wednesday"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

chest

-0.86

åĬ

-0.82

respons

-0.81

abet

-0.78

orius

-0.73

iful

-0.73

Export

-0.73

pend

-0.72

ebook

-0.70

eryl

-0.70

POSITIVE LOGITS

 morning

1.41

 afternoon

1.37

 evening

1.29

 night

1.25

 mornings

1.11

nesday

1.08

 nights

1.05

 evenings

1.03

morning

1.00

 Night

0.95

Activations Density 0.041%

specific days of the week

mentions of the day "Wednesday"

No Comments

No Known Activations