INDEX

Explanations

phrases related to organizations and institutions, especially those associated with civil liberties, healthcare, and politics

oai_token-act-pair · gpt-3.5-turbo

references to civil liberties and associated organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 flats

-0.81

hift

-0.66

ï¸ı

-0.66

GREEN

-0.65

 bleach

-0.64

 irregularities

-0.62

 precincts

-0.60

 Cambodia

-0.60

xus

-0.60

 Palest

-0.58

POSITIVE LOGITS

ICAN

0.90

obile

0.81

 Society

0.77

ilot

0.76

 Institute

0.72

States

0.71

 Association

0.71

 Liberties

0.71

vironment

0.71

 Warehouse

0.68

Activations Density 0.173%

phrases related to organizations and institutions, especially those associated with civil liberties, healthcare, and politics

references to civil liberties and associated organizations

No Comments

No Known Activations