INDEX

Explanations

phrases related to legal and law enforcement activities

oai_token-act-pair · gpt-3.5-turbo

instances of violence or physical altercations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 domestically

-0.78

 curated

-0.73

 sustainability

-0.72

 marquee

-0.70

 fandom

-0.70

 innov

-0.69

 inclusion

-0.69

neb

-0.67

 carrot

-0.67

 aligned

-0.67

POSITIVE LOGITS

Later

1.42

Eventually

1.40

Then

1.39

Upon

1.32

Investigators

1.32

Attempts

1.29

However

1.23

Police

1.22

Meanwhile

1.22

According

1.21

Activations Density 0.381%

phrases related to legal and law enforcement activities

instances of violence or physical altercations

No Comments

No Known Activations