INDEX

Explanations

keywords related to popular culture

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

ural

-0.76

RAW

-0.69

xual

-0.69

thur

-0.69

ERO

-0.67

eki

-0.66

 Centauri

-0.66

 Aviv

-0.64

rets

-0.63

OME

-0.63

POSITIVE LOGITS

ity

1.01

ized

1.01

izing

0.93

ised

0.90

izers

0.88

ization

0.88

izations

0.88

izer

0.82

ize

0.80

popular

0.80

Activations Density 0.537%

keywords related to popular culture

No Comments

No Known Activations