INDEX

Explanations

references to images or visual media

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

dies

-0.74

onite

-0.65

itiz

-0.64

lease

-0.64

merce

-0.63

die

-0.63

laws

-0.63

itent

-0.62

 rapp

-0.62

lies

-0.62

POSITIVE LOGITS

 caption

1.06

Image

0.89

 copyright

0.87

Pic

0.81

 Gallery

0.80

 Â©

0.80

 Images

0.79

çĶ°

0.75

 Photograph

0.75

Images

0.74

Activations Density 0.011%

references to images or visual media

No Comments

No Known Activations