INDEX

Explanations

references to large quantifiable amounts or estimates

oai_token-act-pair · gpt-3.5-turbo

numerical estimates and quantities in discussions of projects or predictions

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

 Cards

-0.51

 Characters

-0.50

 VIDEOS

-0.50

 Spaces

-0.49

 Films

-0.49

 Starcraft

-0.48

âĺħ

-0.46

 Scenes

-0.46

 Streaming

-0.46

 Film

-0.45

POSITIVE LOGITS

 geographically

0.54

 practicable

0.52

 owing

0.49

aterasu

0.48

opted

0.47

 sufficient

0.47

inished

0.47

 jurisd

0.47

unin

0.46

 constitutionally

0.46

Activations Density 4.192%

references to large quantifiable amounts or estimates

numerical estimates and quantities in discussions of projects or predictions

No Comments

No Known Activations