INDEX

Explanations

the symbol '&' and mentions related to specific organizations

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

 mattered

-0.72

etheless

-0.71

owicz

-0.69

ifying

-0.66

orable

-0.65

SPONSORED

-0.64

assian

-0.62

icycle

-0.61

senal

-0.61

 arresting

-0.60

POSITIVE LOGITS

 Friends

0.90

GAN

0.88

Co

0.83

 Associates

0.81

ERSON

0.80

utm

0.79

 Sons

0.78

 Company

0.77

Geh

0.76

optional

0.73

Activations Density 0.417%

the symbol '&' and mentions related to specific organizations

No Comments

No Known Activations