INDEX

Explanations

phrases related to examining and discussing the policies and practices of a company

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

edin

-0.73

mare

-0.72

 Originally

-0.63

venants

-0.62

BR

-0.61

###

-0.60

Â²

-0.60

âĶĢ

-0.59

LAB

-0.59

fleet

-0.59

POSITIVE LOGITS

bidden

1.46

geries

1.37

 example

1.27

gotten

1.26

 instance

1.23

 purposes

1.21

gery

1.20

 starters

1.11

 awhile

1.10

aging

1.09

Activations Density 1.916%

phrases related to examining and discussing the policies and practices of a company

No Comments

No Known Activations