INDEX

Explanations

proper nouns and locations

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

 Sins

-0.71

 Noir

-0.69

 Preferred

-0.65

 Scarlet

-0.63

 Emin

-0.62

 Attribution

-0.61

 Ivory

-0.59

backer

-0.58

 Reson

-0.57

CPC

-0.57

POSITIVE LOGITS

prising

1.32

seless

1.25

pperc

1.17

nexpected

1.15

berman

1.10

pees

1.07

mber

1.07

pee

1.05

gly

1.03

pport

1.01

Activations Density 3.324%

proper nouns and locations

No Comments

No Known Activations