INDEX

Explanations

mentions of ethical matters or breaches

oai_token-act-pair · gpt-3.5-turbo

references to ethics concepts and discussions

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

nces

-0.80

upt

-0.79

 Clockwork

-0.69

xual

-0.69

down

-0.68

 Ingram

-0.67

noon

-0.67

ept

-0.67

Jub

-0.67

eworld

-0.66

POSITIVE LOGITS

onomic

1.11

onom

0.88

 dile

0.83

ostics

0.82

ethical

0.80

 violations

0.76

 watchdog

0.74

 hazard

0.73

 disclosure

0.73

 breaches

0.73

Activations Density 0.029%

mentions of ethical matters or breaches

references to ethics concepts and discussions

No Comments

No Known Activations