INDEX

Explanations

statistics and data related to employment and demographics

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 prosec

-0.88

 oath

-0.86

 sculpt

-0.85

 incarn

-0.81

 enqu

-0.81

 reneg

-0.77

 holog

-0.76

 endeav

-0.76

 himself

-0.74

 lips

-0.74

POSITIVE LOGITS

Overall

1.59

Among

1.53

Compared

1.52

Average

1.52

Percent

1.49

Advertisement

1.47

Despite

1.44

Meanwhile

1.44

Of

1.42

Similarly

1.42

Activations Density 0.389%

statistics and data related to employment and demographics

No Comments

No Known Activations