INDEX

Explanations

words related to political conflicts and military activities

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

 Solitaire

-0.90

Phi

-0.82

 Tags

-0.80

 Compass

-0.76

 Scrib

-0.75

 Jackets

-0.74

 Indigo

-0.70

 Revolution

-0.69

 Shooter

-0.67

 Bravo

-0.67

POSITIVE LOGITS

appointed

1.35

known

1.34

established

1.34

equipped

1.32

trained

1.30

respected

1.30

connected

1.28

defined

1.27

developed

1.26

eyed

1.24

Activations Density 0.355%

words related to political conflicts and military activities

No Comments

No Known Activations