INDEX

Explanations

features and details related to visual elements and descriptions in a document

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

=-=-=-=-=-=-=-=-

-0.79

imester

-0.70

dule

-0.68

iseum

-0.67

 politic

-0.66

ctr

-0.65

 ratification

-0.65

winner

-0.65

reon

-0.65

Honest

-0.65

POSITIVE LOGITS

 protr

1.26

 etched

1.01

 visible

0.97

 indicate

0.97

pots

0.95

 indicating

0.94

 dotted

0.94

 markings

0.94

 emanating

0.93

 denote

0.92

Activations Density 0.133%

features and details related to visual elements and descriptions in a document

No Comments

No Known Activations