INDEX

Explanations

phrases related to criminal activities or police investigations

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

é¾įåĸļå£«

-0.74

merce

-0.71

ura

-0.70

amiya

-0.69

ially

-0.68

taboola

-0.67

ssh

-0.67

oan

-0.66

conn

-0.66

Released

-0.66

POSITIVE LOGITS

 basics

0.83

 whence

0.79

 grid

0.76

 Basics

0.76

GET

0.76

 shore

0.72

 Main

0.69

 Eden

0.69

 Hades

0.69

 Menu

0.69

Activations Density 0.110%

phrases related to criminal activities or police investigations

No Comments

No Known Activations