INDEX

Explanations

articles of clothing, particularly dresses

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ovie

-0.75

ruciating

-0.74

inez

-0.71

 VIDEOS

-0.69

umar

-0.68

 Galile

-0.67

ithmetic

-0.67

auri

-0.66

enthal

-0.66

=-=-=-=-

-0.66

POSITIVE LOGITS

 worn

1.49

 sleeves

1.35

 apparel

1.30

 wearer

1.21

 clothing

1.20

 shirts

1.18

 sleeve

1.18

 shirt

1.17

 garments

1.15

 underwear

1.15

Activations Density 2.489%

articles of clothing, particularly dresses

No Comments

No Known Activations