INDEX

Explanations

phrases related to directions and steps in a process

oai_token-act-pair · gpt-3.5-turbo

phrases indicating significant events or alerts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

actionDate

-0.90

yrim

-0.72

£ı

-0.72

urgical

-0.71

adal

-0.71

maxwell

-0.70

eryl

-0.70

 Emirates

-0.70

 resid

-0.69

 carbohyd

-0.69

POSITIVE LOGITS

Bo

2.34

Bo

2.20

Ho

1.77

bo

1.67

bo

1.67

Ho

1.66

Boo

1.52

BO

1.48

Hoo

1.45

Fo

1.44

Activations Density 0.297%

phrases related to directions and steps in a process

phrases indicating significant events or alerts

No Comments

No Known Activations