INDEX

Explanations

phrases referring to questions or topics of importance

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

aq

-0.78

ugu

-0.64

zig

-0.60

hips

-0.60

hip

-0.58

Bow

-0.57

disk

-0.57

ikuman

-0.57

Bus

-0.56

aneous

-0.56

POSITIVE LOGITS

 ought

1.01

 haun

0.98

 deserves

0.97

 happens

0.97

 arises

0.95

 awaits

0.95

 occurs

0.95

 lends

0.95

 extends

0.94

 consumes

0.94

Activations Density 0.146%

phrases referring to questions or topics of importance

No Comments

No Known Activations