INDEX

Explanations

legal terms and court-related information

oai_token-act-pair · gpt-3.5-turbo

references to legal documents and court-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

hower

-0.70

natureconservancy

-0.68

 buffs

-0.66

isine

-0.65

 selfies

-0.65

 amenities

-0.65

byn

-0.63

 flashes

-0.63

ãĤ©

-0.63

cov

-0.62

POSITIVE LOGITS

 Court

1.50

 court

1.48

 Circuit

1.37

 courts

1.32

 courtroom

1.32

 appellate

1.29

Court

1.28

 tribunal

1.28

 Courts

1.27

 judge

1.26

Activations Density 0.288%

legal terms and court-related information

references to legal documents and court-related terms

No Comments

No Known Activations