INDEX

Explanations

proper nouns and entities such as names, locations, and titles

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

 VIDEOS

-0.89

views

-0.76

inar

-0.71

sticks

-0.69

Characters

-0.67

ophobia

-0.66

aris

-0.66

malink

-0.65

olars

-0.65

estyles

-0.63

POSITIVE LOGITS

 notorious

1.01

 former

1.00

 famed

0.97

 grandson

0.94

 famous

0.90

 nephew

0.89

 infamous

0.86

 deceased

0.85

 charismatic

0.83

 granddaughter

0.83

Activations Density 0.220%

proper nouns and entities such as names, locations, and titles

No Comments

No Known Activations