INDEX

Explanations

information related to politics, military operations, and drone programs

oai_token-act-pair · gpt-3.5-turbo

references to incidents or information involving accusations and military operations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Tickets

-0.77

 chirop

-0.74

 Wilde

-0.72

 ç¥ŀ

-0.71

iour

-0.70

LP

-0.70

 insol

-0.70

club

-0.69

acea

-0.68

ãĤ¦ãĤ¹

-0.68

POSITIVE LOGITS

 counterterrorism

1.48

 Yemeni

1.40

 airstrikes

1.37

 Pentagon

1.36

 airstrike

1.33

 Yemen

1.31

 Iraqi

1.31

 Afghan

1.28

 Taliban

1.27

 Osama

1.26

Activations Density 0.674%

information related to politics, military operations, and drone programs

references to incidents or information involving accusations and military operations

No Comments

No Known Activations