INDEX

Explanations

phrases related to political discussions and events

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

è¦ļéĨĴ

-0.74

 Remastered

-0.72

 cellul

-0.68

arette

-0.67

mat

-0.67

aples

-0.64

ĸļ

-0.63

arettes

-0.61

idated

-0.61

abies

-0.60

POSITIVE LOGITS

eous

1.38

 wing

1.00

 fielder

0.99

wing

0.95

 winger

0.94

 flank

0.85

move

0.84

 handed

0.79

ward

0.79

 smack

0.78

Activations Density 0.325%

phrases related to political discussions and events

No Comments

No Known Activations