INDEX

Explanations

stylistic elements that showcase the manipulation of documents or information

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AND

-0.56

Scene

-0.56

ibrary

-0.54

 Emblem

-0.53

Adds

-0.52

and

-0.52

 today

-0.52

DonaldTrump

-0.50

Warning

-0.50

Filename

-0.50

POSITIVE LOGITS

 thereof

1.20

 alike

1.04

 thereto

1.04

 versa

1.00

 thereafter

0.89

 therein

0.87

 respectively

0.85

cellaneous

0.85

 accordingly

0.83

etc

0.71

Activations Density 6.998%

stylistic elements that showcase the manipulation of documents or information

No Comments

No Known Activations