INDEX

Explanations

any mentions of the word "voting"

oai_token-act-pair · gpt-3.5-turbo

references to voting processes and related activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.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

 Illustrated

-0.74

bane

-0.72

 Archdemon

-0.70

ouf

-0.69

hyde

-0.69

assetsadobe

-0.66

KC

-0.65

tek

-0.62

imb

-0.60

 Monk

-0.59

POSITIVE LOGITS

 voting

3.97

 Voting

3.14

 vote

2.34

 votes

2.22

 ballots

2.06

 voted

2.05

 voter

2.05

 ballot

1.97

 Vote

1.96

Vote

1.94

Activations Density 0.008%

any mentions of the word "voting"

references to voting processes and related activities

No Comments

No Known Activations