INDEX

Explanations

technical references related to computer programming and system performance

oai_token-act-pair · gpt-4o-mini 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

selage

-0.60

 Smithsonian

-0.58

anse

-0.58

ivan

-0.58

Osw

-0.58

itud

-0.58

angered

-0.57

 Urug

-0.57

 Merry

-0.56

HUM

-0.55

POSITIVE LOGITS

 Syndrome

1.00

 syndrome

0.91

 Mode

0.80

 mode

0.80

 shenanigans

0.80

 inhibitors

0.80

atism

0.78

Â®,

0.77

+.

0.76

 techniques

0.76

Activations Density 0.324%

technical references related to computer programming and system performance

No Comments

No Known Activations