INDEX

Explanations

situations or events involving conflict or disruption

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

CI

-0.68

cise

-0.62

selves

-0.60

 ancest

-0.57

lug

-0.57

 whereas

-0.56

dit

-0.54

 Know

-0.53

 Pixel

-0.53

 Sandra

-0.53

POSITIVE LOGITS

 arises

1.16

 goes

1.14

 arose

1.14

 exists

1.12

 explodes

1.11

 flowed

1.11

 disappears

1.10

 becomes

1.10

 begins

1.09

 continues

1.08

Activations Density 2.301%

situations or events involving conflict or disruption

No Comments

No Known Activations