INDEX

Explanations

phrases related to criminal investigations or incidents

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

é¾įåĸļå£«

-0.69

gling

-0.69

edIn

-0.68

merce

-0.68

ographically

-0.67

ssh

-0.67

icipated

-0.66

conn

-0.65

amiya

-0.65

åĩ

-0.65

POSITIVE LOGITS

 Contents

0.86

 Menu

0.82

 Main

0.81

 grid

0.80

 Gallery

0.77

 PAGE

0.77

Top

0.77

bard

0.75

TOP

0.75

 Table

0.73

Activations Density 0.040%

phrases related to criminal investigations or incidents

No Comments

No Known Activations