INDEX

Explanations

proper nouns related to significant places or people

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ãĥĦ

-0.70

é¾įå¥ĳå£«

-0.68

nesota

-0.67

oppable

-0.63

ipeg

-0.62

KP

-0.62

izations

-0.61

 Archdemon

-0.61

ically

-0.60

å£«

-0.59

POSITIVE LOGITS

enne

1.18

uga

1.15

ennes

0.91

amas

0.89

lus

0.87

mans

0.86

apeake

0.86

oning

0.86

ral

0.85

man

0.84

Activations Density 0.003%

proper nouns related to significant places or people

No Comments

No Known Activations