INDEX

Explanations

mentions of negative political situations or criticisms

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

secut

-0.74

CVE

-0.70

 simultane

-0.69

ocent

-0.68

 attentive

-0.66

 uncond

-0.63

SPONSORED

-0.63

 amen

-0.60

etheless

-0.60

 relativity

-0.60

POSITIVE LOGITS

yard

1.06

yards

0.94

shaw

0.93

hire

0.89

books

0.88

book

0.87

aper

0.85

enhagen

0.85

herer

0.83

TING

0.83

Activations Density 0.022%

mentions of negative political situations or criticisms

No Comments

No Known Activations