INDEX

Explanations

numerical values followed by unit measurements, such as kilograms (kg), related to measurements and metrics

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

etheless

-0.75

cknow

-0.74

gotten

-0.73

bidden

-0.73

issance

-0.70

cknowled

-0.68

BOOK

-0.66

Ĥª

-0.66

displayText

-0.64

 Lawn

-0.63

POSITIVE LOGITS

 gorilla

0.95

atsu

0.94

lbs

0.88

emouth

0.86

ross

0.84

iang

0.83

kg

0.83

omez

0.81

 kilograms

0.80

 weight

0.80

Activations Density 0.018%

numerical values followed by unit measurements, such as kilograms (kg), related to measurements and metrics

No Comments

No Known Activations