INDEX

Explanations

statements and quotations attributed to specific individuals

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

 matured

-0.62

 perfected

-0.55

 rigged

-0.53

 pedest

-0.51

 enjoys

-0.50

 assassinate

-0.50

 proved

-0.49

 sleeps

-0.48

 underestimated

-0.47

 experimented

-0.47

POSITIVE LOGITS

 spokesman

0.66

 spokeswoman

0.65

>.

0.63

AFP

0.62

.</

0.58

Xin

0.58

veland

0.57

 Securities

0.56

0.55

soType

0.54

Activations Density 0.530%

statements and quotations attributed to specific individuals

No Comments

No Known Activations