INDEX

Explanations

phrases related to moral concepts

oai_token-act-pair · gpt-3.5-turbo

concepts related to morality and ethics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

xual

-0.79

Lup

-0.70

upon

-0.70

rooms

-0.69

gow

-0.69

rams

-0.69

lers

-0.67

minster

-0.67

hips

-0.66

 Pavilion

-0.66

POSITIVE LOGITS

istic

1.16

izing

1.08

ising

1.05

 compass

1.04

 hazard

1.04

 indignation

1.02

istically

1.00

ised

0.97

IZE

0.95

 dile

0.94

Activations Density 0.033%

phrases related to moral concepts

concepts related to morality and ethics

No Comments

No Known Activations