INDEX

Explanations

names, titles, or keywords related to legal documents

oai_token-act-pair · gpt-3.5-turbo

references to formal agreements, legal documents, and associated terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

aunder

-0.75

earable

-0.71

pez

-0.71

aces

-0.67

age

-0.67

 landfall

-0.66

eah

-0.66

orters

-0.66

iannopoulos

-0.65

orter

-0.65

POSITIVE LOGITS

 Charter

0.92

 Sabha

0.90

 Framework

0.87

izon

0.83

ISM

0.82

kat

0.80

 Amendment

0.78

 Programme

0.74

 Communications

0.73

zman

0.73

Activations Density 0.014%

names, titles, or keywords related to legal documents

references to formal agreements, legal documents, and associated terminology

No Comments

No Known Activations