INDEX

Explanations

names of books, articles, or newspapers, as well as specific terms like 'rights' and 'underwear'

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

fir

-0.78

 scratch

-0.70

 braces

-0.69

 Hoff

-0.68

 drawings

-0.67

 Sawyer

-0.67

WARE

-0.66

 wool

-0.66

 heights

-0.65

 lumber

-0.65

POSITIVE LOGITS

cerpt

1.22

changes

1.21

pelled

1.21

terior

1.18

haust

1.18

ceptions

1.17

posing

1.17

position

1.12

portation

1.12

acts

1.10

Activations Density 1.150%

names of books, articles, or newspapers, as well as specific terms like 'rights' and 'underwear'

No Comments

No Known Activations