INDEX

Explanations

editor's notes in a document

oai_token-act-pair · gpt-3.5-turbo

references to editorial content or notes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Sicily

-0.69

ought

-0.68

ells

-0.67

bley

-0.66

crow

-0.66

 Vulcan

-0.66

pered

-0.65

avid

-0.65

ttp

-0.64

omething

-0.63

POSITIVE LOGITS

ial

0.91

ials

0.87

Editor

0.84

ially

0.80

ror

0.77

ienne

0.76

 Clicker

0.71

editor

0.70

 Picks

0.68

furt

0.68

Activations Density 0.012%

editor's notes in a document

references to editorial content or notes

No Comments

No Known Activations