INDEX

Explanations

technical terms related to computer programming and IT systems

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

encing

-0.72

ASON

-0.71

mble

-0.70

encers

-0.69

 Pradesh

-0.67

pter

-0.64

enced

-0.63

encer

-0.62

LV

-0.61

aeda

-0.61

POSITIVE LOGITS

strap

1.23

wana

1.00

jamin

0.97

zeb

0.92

aign

0.92

cies

0.91

yond

0.88

arest

0.88

jee

0.85

ghazi

0.83

Activations Density 4.924%

technical terms related to computer programming and IT systems

No Comments

No Known Activations