INDEX

Explanations

phrases related to legal frameworks or codes

oai_token-act-pair · gpt-3.5-turbo

references to legal documents or statutes

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

 blame

-0.72

 lingering

-0.71

 staged

-0.67

 snowball

-0.67

 drops

-0.66

 skewed

-0.66

 cheaply

-0.65

ze

-0.65

 tipping

-0.65

 chem

-0.65

POSITIVE LOGITS

shall

1.48

Marginal

1.41

NRS

1.40

CHAPTER

1.38

Section

1.36

 Notwithstanding

1.36

Â§

1.34

Exception

1.28

Definition

1.26

 Definitions

1.24

Activations Density 0.164%

phrases related to legal frameworks or codes

references to legal documents or statutes

No Comments

No Known Activations