INDEX

Explanations

proper nouns relating to locations and franchises

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

-0.76

mble

-0.76

 timet

-0.71

 Prol

-0.62

 Aber

-0.60

 exerc

-0.58

 intrinsic

-0.56

berra

-0.56

 Tiny

-0.55

 guiName

-0.55

POSITIVE LOGITS

uterte

0.69

Tex

0.68

addafi

0.67

hovah

0.64

 Islanders

0.63

ropolis

0.62

 Vegas

0.61

 Janeiro

0.61

 Colo

0.60

 Mexicans

0.60

Activations Density 1.801%

proper nouns relating to locations and franchises

No Comments

No Known Activations