INDEX

Explanations

topics or subjects that are being highlighted or brought to attention by shedding light on them

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asus

-0.79

 nomine

-0.73

ĳå£«

-0.67

 Rampage

-0.66

ocide

-0.63

 renegoti

-0.62

apons

-0.62

 recip

-0.62

ction

-0.61

hare

-0.60

POSITIVE LOGITS

lights

0.98

enment

0.98

nings

0.97

 illuminating

0.93

 illuminate

0.92

 shone

0.87

eners

0.86

hearted

0.85

 clues

0.84

ening

0.83

Activations Density 0.024%

topics or subjects that are being highlighted or brought to attention by shedding light on them

No Comments

No Known Activations