INDEX

Explanations

words related to the term 'angel'

oai_token-act-pair · gpt-3.5-turbo

instances of the substring "ang"

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nesday

-0.71

 compensated

-0.69

ãĥīãĥ©

-0.68

Ö¼

-0.68

ittee

-0.67

ngth

-0.66

cloth

-0.65

parable

-0.64

Pwr

-0.64

ometimes

-0.64

POSITIVE LOGITS

aroo

1.26

omez

1.00

etsu

0.97

river

0.92

alore

0.90

zhou

0.88

lia

0.88

irl

0.87

ethe

0.87

regate

0.86

Activations Density 0.027%

words related to the term 'angel'

instances of the substring "ang"

No Comments

No Known Activations