INDEX

Explanations

the term "actual" or variations of it in different contexts

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

 Beware

-0.75

wich

-0.75

zy

-0.69

 Azerb

-0.68

nan

-0.66

Gate

-0.63

 surely

-0.63

ervative

-0.62

limit

-0.62

 Vaugh

-0.62

POSITIVE LOGITS

ity

1.00

isation

0.99

izations

0.98

izable

0.97

ities

0.91

ignment

0.88

isations

0.88

idad

0.86

ITY

0.85

 malice

0.82

Activations Density 0.015%

the term "actual" or variations of it in different contexts

No Comments

No Known Activations