INDEX

Explanations

phrases related to political events and statements

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

-0.69

pload

-0.69

 Bowl

-0.64

angs

-0.63

leaf

-0.62

 Stard

-0.60

roups

-0.59

bole

-0.57

lez

-0.56

 Harbor

-0.56

POSITIVE LOGITS

 affairs

0.92

Ĥİ

0.73

 Decay

0.70

hyde

0.69

vironment

0.68

azeera

0.68

inence

0.67

CLASSIFIED

0.66

usted

0.65

 emergency

0.64

Activations Density 0.876%

phrases related to political events and statements

No Comments

No Known Activations