INDEX

Explanations

praise or criticism in reviews about various forms of art such as movies, music, and television shows

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

 reimbursement

-0.86

 credential

-0.85

 Certification

-0.84

Fax

-0.84

 receipt

-0.80

Federal

-0.80

iership

-0.78

 revoked

-0.75

 Member

-0.75

IRC

-0.75

POSITIVE LOGITS

 themes

1.36

lyr

1.35

 juxtap

1.31

 genres

1.27

 beautifully

1.26

 cinematic

1.26

 visuals

1.23

 genre

1.20

 storytelling

1.19

 textures

1.18

Activations Density 5.538%

praise or criticism in reviews about various forms of art such as movies, music, and television shows

No Comments

No Known Activations