INDEX

Explanations

sources or credits attributions in text

oai_token-act-pair · gpt-3.5-turbo

references to sources or attribution in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

adal

-0.83

inary

-0.77

lesh

-0.71

ecause

-0.71

isol

-0.69

thin

-0.68

ties

-0.68

 accordingly

-0.68

rounder

-0.65

thy

-0.63

POSITIVE LOGITS

 Wikimedia

1.27

 Shutterstock

1.16

 Flickr

1.03

 Getty

0.99

 Yelp

0.96

 REUTERS

0.94

 Huffington

0.92

 Wikipedia

0.89

 Twitter

0.88

 Facebook

0.87

Activations Density 0.023%

sources or credits attributions in text

references to sources or attribution in text

No Comments

No Known Activations