INDEX

Explanations

words related to important or significant concepts or ideas

oai_token-act-pair · gpt-3.5-turbo

important concepts or significant elements within a text

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

ews

-0.71

 Tsukuyomi

-0.71

uthor

-0.68

asca

-0.67

AUT

-0.67

 Bett

-0.67

 Horses

-0.66

 Mostly

-0.66

 Chatt

-0.66

 Fever

-0.65

POSITIVE LOGITS

stone

1.19

stroke

1.19

stro

1.01

stones

0.99

*/(

0.90

binding

0.90

wcs

0.89

hole

0.89

ring

0.88

key

0.86

Activations Density 0.020%

words related to important or significant concepts or ideas

important concepts or significant elements within a text

No Comments

No Known Activations