INDEX

Explanations

phrases related to politics and government, specifically mentioning names of politicians, government officials, and political events

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

FTP

-0.68

SAM

-0.68

NAD

-0.67

iencies

-0.67

ASA

-0.67

NAT

-0.64

ECD

-0.63

 mater

-0.62

BILITY

-0.61

 Emacs

-0.60

POSITIVE LOGITS

care

0.97

eting

0.88

TRUMP

0.87

Trump

0.86

eters

0.83

 Donald

0.81

 Trump

0.81

eter

0.81

 supporters

0.80

Donald

0.79

Activations Density 1.472%

phrases related to politics and government, specifically mentioning names of politicians, government officials, and political events

No Comments

No Known Activations