INDEX

Explanations

phrases related to legal complaints and documents

oai_token-act-pair · gpt-3.5-turbo

negative statements or phrases that express a lack of validation or certainty

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

 knows

-0.66

 keeps

-0.64

 waits

-0.64

 likes

-0.63

 hopping

-0.63

 loves

-0.62

 dying

-0.62

TPPStreamerBot

-0.60

 invincible

-0.60

 sleeps

-0.59

POSITIVE LOGITS

 involve

1.66

 relate

1.34

 entail

1.30

 originate

1.29

 contain

1.25

 consist

1.24

 hinge

1.23

 encompass

1.23

 reflect

1.20

 stem

1.17

Activations Density 0.368%

phrases related to legal complaints and documents

negative statements or phrases that express a lack of validation or certainty

No Comments

No Known Activations