INDEX

Explanations

references to political leaders and positions in government

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

anwhile

-0.76

oldemort

-0.74

 Canaver

-0.70

ascade

-0.67

INESS

-0.67

afe

-0.66

oteric

-0.66

 Humanity

-0.65

okia

-0.63

 Beir

-0.63

POSITIVE LOGITS

 ministerial

1.17

 minister

1.11

 Minister

1.04

knit

0.98

 ministers

0.95

val

0.94

 Ministers

0.84

etime

0.80

iary

0.78

ises

0.76

Activations Density 0.420%

references to political leaders and positions in government

No Comments

No Known Activations