INDEX

Explanations

words related to politics and elections, focusing on swing seats, majorities, and party support

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.91

obyl

-0.75

ipedia

-0.74

ICLE

-0.72

assetsadobe

-0.69

Technology

-0.67

uador

-0.66

ibaba

-0.66

llah

-0.65

SOURCE

-0.65

POSITIVE LOGITS

 constituencies

1.12

 districts

1.05

 Elections

1.00

 elections

0.99

 constituency

0.98

holder

0.98

 vacated

0.97

holders

0.96

 votes

0.94

 electorate

0.92

Activations Density 0.098%

words related to politics and elections, focusing on swing seats, majorities, and party support

No Comments

No Known Activations