INDEX

Explanations

instances where the word "and" appears

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ibly

-0.86

uously

-0.83

icity

-0.80

ious

-0.78

 unfocusedRange

-0.73

utenberg

-0.71

asar

-0.67

isoft

-0.65

ALLY

-0.64

blogspot

-0.62

POSITIVE LOGITS

romeda

0.66

 ears

0.63

bern

0.62

 feet

0.62

 sisters

0.62

 shoulders

0.61

 unic

0.61

 bolts

0.60

 grandparents

0.59

asus

0.58

Activations Density 0.205%

instances where the word "and" appears

No Comments

No Known Activations