INDEX

Explanations

phrases related to the concept of pieces of a whole or parts of something

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

few

-0.74

parts

-0.74

heses

-0.72

abilities

-0.70

ombies

-0.70

days

-0.70

terms

-0.69

most

-0.69

ivals

-0.68

isodes

-0.68

POSITIVE LOGITS

 cloth

0.81

 sunshine

0.80

 luck

0.73

gum

0.73

 ammunition

0.72

 cash

0.70

 doom

0.70

 asphalt

0.69

 extra

0.68

 silver

0.67

Activations Density 0.766%

phrases related to the concept of pieces of a whole or parts of something

No Comments

No Known Activations