INDEX

Explanations

references to images in a document

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eller

-0.92

gew

-0.85

iated

-0.82

indal

-0.80

iates

-0.80

etermination

-0.77

iations

-0.75

âĵĺ

-0.74

ose

-0.73

iating

-0.71

POSITIVE LOGITS

 IMAGES

0.86

 courtesy

0.79

 Features

0.73

 Thumbnails

0.72

 Shutterstock

0.72

lihood

0.71

upload

0.71

cour

0.70

clips

0.70

Ø³

0.70

Activations Density 0.020%

references to images in a document

No Comments

No Known Activations