INDEX

Explanations

dates mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "date" and variations of it

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

HAEL

-0.83

BLIC

-0.80

raq

-0.77

estern

-0.76

æĸ¹

-0.76

uyomi

-0.75

luaj

-0.73

IDES

-0.73

olulu

-0.72

 Flavoring

-0.71

POSITIVE LOGITS

TBD

0.78

 dates

0.73

 stamp

0.71

 inacc

0.70

 stamps

0.69

 rape

0.68

 date

0.68

dates

0.68

lled

0.67

days

0.65

Activations Density 0.030%

dates mentioned in the text

instances of the word "date" and variations of it

No Comments

No Known Activations