INDEX

Explanations

mentions of popular culture references, politics, and appointments

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

ï¸ı

-0.73

axter

-0.69

*/(

-0.65

 injection

-0.63

halla

-0.63

PsyNetMessage

-0.60

 bleach

-0.59

ittal

-0.59

iasco

-0.58

ividual

-0.58

POSITIVE LOGITS

ĨĴ

0.84

å£«

0.69

 thous

0.68

 Warcraft

0.67

bia

0.64

eteenth

0.63

 Thrones

0.63

SEA

0.63

merce

0.60

 Emirates

0.58

Activations Density 2.362%

mentions of popular culture references, politics, and appointments

No Comments

No Known Activations