INDEX

Explanations

mentions of political figures and positions, mainly related to the United States government and Congress

oai_token-act-pair · gpt-3.5-turbo

references to political positions and governmental institutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

plet

-0.64

kinson

-0.63

 boxed

-0.61

forms

-0.61

 maneuver

-0.60

gra

-0.60

 penetrating

-0.57

 boxes

-0.57

plets

-0.57

cover

-0.57

POSITIVE LOGITS

 Sabha

0.83

inence

0.69

assad

0.67

wark

0.66

itaire

0.66

 overseeing

0.65

thood

0.63

otonin

0.63

Cola

0.62

Lia

0.62

Activations Density 0.208%

mentions of political figures and positions, mainly related to the United States government and Congress

references to political positions and governmental institutions

No Comments

No Known Activations