INDEX

Explanations

phrases related to news and reports, specifically those discussing medical systems and practices

oai_token-act-pair · gpt-3.5-turbo

queries related to information gathering or inquiries about specific topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 estab

-0.73

 himself

-0.70

 democrat

-0.68

 ridic

-0.67

 rightful

-0.66

 dismant

-0.65

 sculpt

-0.65

 pacif

-0.65

 appe

-0.64

 renegoti

-0.64

POSITIVE LOGITS

Guest

1.13

Contents

1.10

Introduction

1.03

Overview

1.03

Description

0.99

Welcome

0.99

Published

0.99

Details

0.98

Whether

0.98

NAS

0.96

Activations Density 0.231%

phrases related to news and reports, specifically those discussing medical systems and practices

queries related to information gathering or inquiries about specific topics

No Comments

No Known Activations