INDEX

Explanations

the word "Mate" in the given text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

icum

-0.79

icity

-0.78

acea

-0.74

lich

-0.73

shire

-0.72

kefeller

-0.69

ights

-0.68

acco

-0.66

lander

-0.65

icals

-0.65

POSITIVE LOGITS

vich

0.86

 Mate

0.85

 mate

0.84

volent

0.80

 mates

0.78

 versa

0.76

creen

0.72

asus

0.72

thur

0.71

llan

0.71

Activations Density 0.024%

the word "Mate" in the given text

No Comments

No Known Activations