INDEX

Explanations

dialogue or quotation marks surrounding speech

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 standardized

-0.82

 denomin

-0.81

 tariff

-0.81

 manufacturer

-0.80

 tablet

-0.79

 licens

-0.78

 assigned

-0.78

 endors

-0.78

 wage

-0.76

 administr

-0.75

POSITIVE LOGITS

 Dirty

1.24

 Panic

1.22

 Goodbye

1.21

 Revenge

1.20

 Haunted

1.20

 Fever

1.20

 Sleeping

1.19

 Trouble

1.16

 Falling

1.15

 Salvation

1.15

Activations Density 0.103%

dialogue or quotation marks surrounding speech

No Comments

No Known Activations