INDEX

Explanations

phrases related to specific entities or organizations, such as names of people, universities, and products

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

sburgh

-0.83

ulators

-0.81

ulatory

-0.70

hell

-0.69

fram

-0.69

ships

-0.68

âĸĪâĸĪ

-0.68

ambo

-0.67

ocene

-0.67

akia

-0.66

POSITIVE LOGITS

ynamic

0.88

cember

0.83

ependent

0.78

ynam

0.77

hyde

0.77

Mub

0.74

etermination

0.73

iscovery

0.72

etermin

0.71

NF

0.69

Activations Density 0.340%

phrases related to specific entities or organizations, such as names of people, universities, and products

No Comments

No Known Activations