INDEX

Explanations

keywords related to parts of a series or sequence, such as "installment" and "iteration."

oai_token-act-pair · gpt-3.5-turbo

references to installments or iterations in media and series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

eric

-0.80

gans

-0.79

yer

-0.79

*/(

-0.76

inates

-0.72

icial

-0.69

bara

-0.69

eating

-0.69

inators

-0.68

mouth

-0.68

POSITIVE LOGITS

 installment

1.18

 installments

1.00

ĸļ

0.86

itial

0.75

VII

0.74

 VIII

0.69

teenth

0.69

XII

0.68

DonaldTrump

0.67

 iteration

0.67

Activations Density 0.012%

keywords related to parts of a series or sequence, such as "installment" and "iteration."

references to installments or iterations in media and series

No Comments

No Known Activations