INDEX

Explanations

phrases related to cause and effect or prediction

oai_token-act-pair · gpt-3.5-turbo

the word "just" and its contexts in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 challeng

-0.75

glomer

-0.67

 thereof

-0.66

existent

-0.66

 cous

-0.65

senal

-0.64

conservancy

-0.62

 acquaint

-0.62

externalActionCode

-0.62

 conflic

-0.62

POSITIVE LOGITS

ices

1.18

ifications

1.17

itia

1.02

ifiable

1.00

icia

0.91

ified

0.91

if

0.90

ification

0.88

Giving

0.87

ifi

0.86

Activations Density 0.060%

phrases related to cause and effect or prediction

the word "just" and its contexts in the text

No Comments

No Known Activations