INDEX

Explanations

instances where instructions are given or steps are listed

oai_token-act-pair · gpt-3.5-turbo

references to specific indicators or components in a technical context

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

SPONSORED

-1.02

âĢİ

-0.79

 Maher

-0.74

AFP

-0.73

 Tasmania

-0.72

Malley

-0.71

-0.69

Catholic

-0.69

CTV

-0.68

AIDS

-0.68

POSITIVE LOGITS

 same

1.26

 entire

1.25

 remainder

1.24

ses

1.21

 simplest

1.20

 aforementioned

1.18

 entirety

1.18

 easiest

1.17

 widest

1.14

 lowest

1.12

Activations Density 1.143%

instances where instructions are given or steps are listed

references to specific indicators or components in a technical context

No Comments

No Known Activations