INDEX

Explanations

terms related to political establishments

oai_token-act-pair · gpt-3.5-turbo

references to political or organizational systems of authority

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Marketable

-0.83

tnc

-0.79

range

-0.77

brid

-0.74

cific

-0.74

ming

-0.69

ranged

-0.68

ioxide

-0.68

 Emin

-0.67

spell

-0.67

POSITIVE LOGITS

ablishment

1.20

 establishment

1.02

 Establishment

0.96

 establishments

0.93

 conservatism

0.82

 affili

0.81

 elites

0.79

itutes

0.78

establishment

0.76

 hierarch

0.75

Activations Density 0.013%

terms related to political establishments

references to political or organizational systems of authority

No Comments

No Known Activations