INDEX

Explanations

phrases related to technology and government regulation

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

FTWARE

-0.60

igers

-0.59

 Dragons

-0.59

 Fired

-0.59

Leaks

-0.57

olutions

-0.57

 Releases

-0.57

aukee

-0.56

 WARRANT

-0.54

izon

-0.54

POSITIVE LOGITS

 alas

1.01

uh

0.97

 unsurprisingly

0.93

 somew

0.88

um

0.88

cru

0.85

 frankly

0.84

 controvers

0.83

 importantly

0.82

 ironically

0.80

Activations Density 3.718%

phrases related to technology and government regulation

No Comments

No Known Activations