INDEX

Explanations

phrases with unique characters like "Ã§" and "ÅŁ"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the character "ç."

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

 Breed

-0.71

 mileage

-0.67

 benefic

-0.67

 rooting

-0.66

 mutual

-0.65

 trail

-0.63

 compuls

-0.62

 marrow

-0.62

OSP

-0.61

ACTED

-0.60

POSITIVE LOGITS

Ã£o

1.25

Ã§

1.06

Ã£

1.03

uration

0.96

arta

0.96

onne

0.94

Ã¡n

0.91

Ã¼

0.88

ĩ

0.88

Ã³n

0.87

Activations Density 0.006%

phrases with unique characters like "Ã§" and "ÅŁ"

occurrences of the character "ç."

No Comments

No Known Activations