INDEX

Explanations

technical instructions or guides

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

 Mobil

-0.69

 Mirage

-0.66

Xan

-0.64

 Antar

-0.64

 mosqu

-0.63

 vulner

-0.63

 Manit

-0.63

 seiz

-0.63

 notebooks

-0.62

iling

-0.62

POSITIVE LOGITS

ï¸ı

0.95

should

0.93

shall

0.92

Pg

0.91

 denotes

0.88

could

0.86

 belongs

0.84

felt

0.83

udes

0.83

 deserves

0.82

Activations Density 0.822%

technical instructions or guides

No Comments

No Known Activations