INDEX

Explanations

phrases related to a specific entity or person, specifically "Pan" and variations like "Panetta"

oai_token-act-pair · gpt-3.5-turbo

references to "Pan" or "Pan American."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ãĤī

-0.68

éĩ

-0.68

ãģĦ

-0.65

ĳå£«

-0.64

ENC

-0.64

çĦ

-0.64

 guiIcon

-0.64

 Gilmore

-0.63

é¾įå¥ĳå£«

-0.63

ï¸ı

-0.63

POSITIVE LOGITS

orama

1.22

omaly

1.00

alyst

0.99

tera

0.97

ahon

0.97

Pan

0.95

acea

0.92

amera

0.90

zers

0.90

agos

0.90

Activations Density 0.012%

phrases related to a specific entity or person, specifically "Pan" and variations like "Panetta"

references to "Pan" or "Pan American."

No Comments

No Known Activations