INDEX

Explanations

phrases indicating fear for someone's safety or life

oai_token-act-pair · gpt-3.5-turbo

concerns related to fear for personal safety or life

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

•layer 8 - fear

No Comments

Negative Logits

AMA

-0.57

 Americans

-0.55

 humans

-0.54

 tourists

-0.52

 gays

-0.52

 americ

-0.50

 African

-0.49

 American

-0.49

 consumers

-0.49

 locals

-0.48

POSITIVE LOGITS

 accomp

0.68

 Citiz

0.68

 conclud

0.64

idable

0.60

 occas

0.58

atus

0.57

Move

0.57

 ende

0.57

ä¹ĭ

0.56

æĪ¦

0.55

Activations Density 0.668%

phrases indicating fear for someone's safety or life

concerns related to fear for personal safety or life

No Comments

No Known Activations