INDEX

Explanations

compound words or phrases that include a specific keyword

oai_token-act-pair · gpt-3.5-turbo

terms related to work and professional environments

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 breathing

-0.68

 penn

-0.66

EGA

-0.65

agnar

-0.64

rha

-0.63

 clen

-0.63

iatrics

-0.63

pling

-0.62

INST

-0.59

lap

-0.59

POSITIVE LOGITS

er

1.97

ership

1.62

ers

1.55

erd

1.23

erness

1.15

erate

1.08

eric

1.01

eri

1.00

ation

0.95

ER

0.93

Activations Density 0.054%

compound words or phrases that include a specific keyword

terms related to work and professional environments

No Comments

No Known Activations