INDEX

Explanations

references to specific terms and jargon from computer programming and technology

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

assetsadobe

-0.76

-0.67

[];

-0.66

"\

-0.63

"<

-0.63

 obser

-0.61

jun

-0.61

 è£ıè

-0.60

Gle

-0.60

 à¨

-0.59

POSITIVE LOGITS

eware

0.76

raid

0.72

ax

0.72

ormon

0.71

zon

0.70

tex

0.69

ipers

0.69

rake

0.69

act

0.69

eez

0.68

Activations Density 2.264%

references to specific terms and jargon from computer programming and technology

No Comments

No Known Activations