INDEX

Explanations

references to legal terms, such as sections of laws or codes

oai_token-act-pair · gpt-3.5-turbo

references to legal articles, sections, and clauses

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 dots

-0.77

 doubts

-0.74

artifacts

-0.69

 distractions

-0.67

 fooled

-0.67

 firsthand

-0.66

 immersed

-0.65

aturdays

-0.65

 mistakes

-0.64

 glimps

-0.64

POSITIVE LOGITS

 Directive

0.90

 Regulations

0.88

 Regulation

0.86

 Annex

0.85

 Schedule

0.83

 Protocol

0.79

NRS

0.79

 Code

0.76

 Constitution

0.76

 Appendix

0.73

Activations Density 0.075%

references to legal terms, such as sections of laws or codes

references to legal articles, sections, and clauses

No Comments

No Known Activations