INDEX

Explanations

percentages and numerical values

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Rav

-0.71

kefeller

-0.64

Attributes

-0.63

flies

-0.62

ITED

-0.62

iden

-0.60

pload

-0.57

Merit

-0.57

ãĤ©

-0.56

hran

-0.56

POSITIVE LOGITS

ABV

0.93

xual

0.88

imet

0.80

iles

0.79

+.

0.79

+)

0.78

 utilization

0.77

imeter

0.75

 humidity

0.73

uary

0.70

Activations Density 0.829%

percentages and numerical values

No Comments

No Known Activations