INDEX

Explanations

mention of political events and figures, particularly related to investigations and denials

oai_token-act-pair · gpt-3.5-turbo

references to official statements or denials related to political events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

urban

-0.89

 ILCS

-0.88

 Auckland

-0.88

 glim

-0.85

 Jagu

-0.83

 herds

-0.81

 Totem

-0.81

Urban

-0.80

tnc

-0.80

iculture

-0.79

POSITIVE LOGITS

FBI

1.51

 Comey

1.48

FBI

1.41

 WikiLeaks

1.32

DOJ

1.29

 Clapper

1.29

 subpoen

1.27

Leaks

1.23

 Rosenstein

1.20

 Benghazi

1.19

Activations Density 1.076%

mention of political events and figures, particularly related to investigations and denials

references to official statements or denials related to political events

No Comments

No Known Activations