INDEX

Explanations

terminology related to teamwork and group efforts

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

 Fountain

-0.68

alam

-0.66

ãĥī

-0.66

tml

-0.63

piration

-0.60

OPLE

-0.60

Consumer

-0.60

DonaldTrump

-0.58

 Godd

-0.58

Maw

-0.58

POSITIVE LOGITS

members

1.64

 members

1.50

mates

1.37

 member

1.36

member

1.36

 mates

1.25

rons

1.17

Members

1.17

 disbanded

1.16

 Members

1.14

Activations Density 4.273%

terminology related to teamwork and group efforts

No Comments

No Known Activations