INDEX

Explanations

the word "normal" or variations of it

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "normal."

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

iosyncr

-0.77

hani

-0.76

haw

-0.72

Lens

-0.70

hop

-0.70

NRS

-0.69

Sov

-0.69

Spot

-0.68

artisan

-0.68

otle

-0.64

POSITIVE LOGITS

ization

1.18

isation

1.18

cy

1.16

izes

1.16

ised

1.12

izing

1.09

ises

1.05

ize

1.05

izers

1.01

ized

0.99

Activations Density 0.024%

the word "normal" or variations of it

references to the concept of "normal."

No Comments

No Known Activations