INDEX

Explanations

phrases related to direction or positioning, specifically involving the concept of "back."

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "back-and-forth" or concepts related to reciprocation and repetition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Parenthood

-0.69

ulhu

-0.67

ãĤ¼ãĤ¦ãĤ¹

-0.63

ãģ®éŃĶ

-0.62

 mmol

-0.61

Phi

-0.60

Duo

-0.60

 Caucasus

-0.57

»Ĵ

-0.56

IX

-0.56

POSITIVE LOGITS

arching

1.11

to

1.07

office

1.07

end

1.05

story

1.00

door

1.00

up

0.99

run

0.98

off

0.97

of

0.95

Activations Density 0.059%

phrases related to direction or positioning, specifically involving the concept of "back."

instances of the phrase "back-and-forth" or concepts related to reciprocation and repetition

No Comments

No Known Activations