INDEX

Explanations

phrases related to political actions and policies

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

 partName

-0.75

Spoiler

-0.72

agonist

-0.71

tick

-0.69

 EVENT

-0.65

agents

-0.64

DragonMagazine

-0.63

perfect

-0.63

ensional

-0.63

================================

-0.63

POSITIVE LOGITS

 subsidies

1.03

 privat

0.97

 amnesty

0.94

 repatri

0.91

 tariffs

0.88

 offshore

0.86

 deport

0.86

 privatization

0.86

 vouchers

0.85

 pensions

0.85

Activations Density 0.530%

phrases related to political actions and policies

No Comments

No Known Activations