INDEX

Explanations

names or terms with accented characters

oai_token-act-pair · gpt-3.5-turbo

occurrences of a specific character or symbol

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

uyomi

-0.79

NetMessage

-0.73

Downloadha

-0.70

insula

-0.66

hole

-0.65

Zip

-0.63

 Decay

-0.63

 surgeons

-0.62

 buckle

-0.62

 crush

-0.60

POSITIVE LOGITS

ÃŃ

1.25

Ã³n

1.02

Ã°

0.94

Ã£o

0.92

Ãį

0.90

Ã¡n

0.89

Ã¡

0.88

Å

0.88

Ãī

0.88

ÃŃn

0.88

Activations Density 0.005%

names or terms with accented characters

occurrences of a specific character or symbol

No Comments

No Known Activations