INDEX

Explanations

news stories that are considered important or significant

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

ogly

-0.77

tails

-0.73

 nurs

-0.72

 toget

-0.71

abase

-0.71

 ener

-0.71

ãĤ¦ãĤ¹

-0.69

 independ

-0.68

 challeng

-0.66

ĪĴ

-0.66

POSITIVE LOGITS

 Thousands

0.81

 Researchers

0.77

 Surveillance

0.77

ï

0.76

 Authorities

0.75

 Investigators

0.75

 Latest

0.74

 Kaine

0.74

 Highlights

0.74

 Macron

0.73

Activations Density 0.010%

news stories that are considered important or significant

No Comments

No Known Activations

news stories that are considered important or significant

No Comments

No Known Activations