INDEX

Explanations

names of people from various backgrounds and professions

oai_token-act-pair · gpt-3.5-turbo

names of individuals along with their professions or roles

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

ials

-0.71

LESS

-0.66

 tides

-0.66

antics

-0.65

lyn

-0.65

rent

-0.63

 Stories

-0.63

rington

-0.63

Ħ¢

-0.63

~~~~~~~~

-0.62

POSITIVE LOGITS

 deputy

1.41

 spokesperson

1.26

 secretary

1.22

 coordinator

1.22

 chairman

1.21

 spokesman

1.20

 lawmaker

1.17

 director

1.10

 Deputy

1.10

 head

1.10

Activations Density 0.153%

names of people from various backgrounds and professions

names of individuals along with their professions or roles

No Comments

No Known Activations