INDEX

Explanations

items listed in a series format preceded by a colon

oai_token-act-pair · gpt-3.5-turbo

phrases related to components and their interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 unethical

-0.76

 disillusion

-0.74

 atheists

-0.72

 desper

-0.72

 mete

-0.71

assetsadobe

-0.69

 harsher

-0.67

 cynical

-0.66

 cynicism

-0.65

 unthinkable

-0.65

POSITIVE LOGITS

 contents

0.97

 exterior

0.95

 subsystem

0.94

 nucleus

0.92

 main

0.90

 Secondary

0.90

 folder

0.89

 sections

0.88

 Header

0.87

 components

0.86

Activations Density 0.591%

items listed in a series format preceded by a colon

phrases related to components and their interactions

No Comments

No Known Activations