On 2/4/25 3:06 AM, Mild Shock wrote:
Hi,
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Group Preference Optimization
https://arxiv.org/abs/2310.11523
DeepSeekMath: Pushing the Limits
https://arxiv.org/abs/2402.03300
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
Bye
Mild Shock schrieb:
Hi,
Wait till USA figures out there is a second
competitor besides DeepSeek, its called Yi-Lightning:
Yi-Lightning Technical Report
https://arxiv.org/abs/2412.01253
It was already discussed 2 months ago:
Eric Schmidt DROPS BOMBSHELL: China DOMINATES AI!
https://www.youtube.com/watch?v=ddWuEUjo4u4
Bye
Throwing papers at us (i.e. in usenet) although better than nothing, it
still is an extremely inefficient way to communicate.
Speak, yourself, in the way that a first year undergraduate prof
explains stuff to the students who just a few months earlier finished
their high schooling.
Then you will communicate.
If it is too difficult to do that (cause requires deep understanding of
the subject by yourself), or, if you don't get time doing all that, then break the job down into little pieces. A day at a time, or a week at a
time.
You'd be spending the same amount of time as you are now, and yet you
will actually communicate something with those with physics background.
Did I have to say this? I don't think so. But your intentions here are
vague. But if you need support for your own gains in understanding the subject, you just have to do it the way I explained above. Then you will
get that support. Not from Bozos and "engineers", but from me and
anybody with actual physics background and such trainings.
There are a few actual physicists in relativity group that can join in,
but only if you do it the way I explained above.
to generate now thinking
to generate new thinking
Hi,
You wouldn't notic that I speak for myself
always and ever. Just remove the links and
you see its all text original by me.
I always post geniun commentary to current
events which are solely the product of my
own mind. I do not intend to educate something.
My intention is to document my current
understanding and to generate now thinking.
If you have expected some schooling, I am very
sorry for you. I never post boring schooling.
Bye
---------------- cut here -----------------
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
---------------- cut here -----------------
Physfitfreak schrieb:
On 2/4/25 3:06 AM, Mild Shock wrote:
Hi,
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Group Preference Optimization
https://arxiv.org/abs/2310.11523
DeepSeekMath: Pushing the Limits
https://arxiv.org/abs/2402.03300
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
Bye
Mild Shock schrieb:
Hi,
Wait till USA figures out there is a second
competitor besides DeepSeek, its called Yi-Lightning:
Yi-Lightning Technical Report
https://arxiv.org/abs/2412.01253
It was already discussed 2 months ago:
Eric Schmidt DROPS BOMBSHELL: China DOMINATES AI!
https://www.youtube.com/watch?v=ddWuEUjo4u4
Bye
Throwing papers at us (i.e. in usenet) although better than nothing,
it still is an extremely inefficient way to communicate.
Speak, yourself, in the way that a first year undergraduate prof
explains stuff to the students who just a few months earlier finished
their high schooling.
Then you will communicate.
If it is too difficult to do that (cause requires deep understanding
of the subject by yourself), or, if you don't get time doing all that,
then break the job down into little pieces. A day at a time, or a week
at a time.
You'd be spending the same amount of time as you are now, and yet you
will actually communicate something with those with physics background.
Did I have to say this? I don't think so. But your intentions here are
vague. But if you need support for your own gains in understanding the
subject, you just have to do it the way I explained above. Then you
will get that support. Not from Bozos and "engineers", but from me and
anybody with actual physics background and such trainings.
There are a few actual physicists in relativity group that can join
in, but only if you do it the way I explained above.
Group Preference Optimization
https://arxiv.org/abs/2310.11523
DeepSeekMath: Pushing the Limits
https://arxiv.org/abs/2402.03300
Hi,
Typo: now -> new
to generate now thinking
Should read:
to generate new thinking
I shit and piss on some asslicking group
that cannot think on his own. Thank god
USENET has only rarely such group morons.
Bye
Mild Shock schrieb:
Hi,
You wouldn't notic that I speak for myself
always and ever. Just remove the links and
you see its all text original by me.
I always post geniun commentary to current
events which are solely the product of my
own mind. I do not intend to educate something.
My intention is to document my current
understanding and to generate now thinking.
If you have expected some schooling, I am very
sorry for you. I never post boring schooling.
Bye
---------------- cut here -----------------
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
---------------- cut here -----------------
Physfitfreak schrieb:
On 2/4/25 3:06 AM, Mild Shock wrote:
Hi,
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Group Preference Optimization
https://arxiv.org/abs/2310.11523
DeepSeekMath: Pushing the Limits
https://arxiv.org/abs/2402.03300
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
Bye
Mild Shock schrieb:
Hi,
Wait till USA figures out there is a second
competitor besides DeepSeek, its called Yi-Lightning:
Yi-Lightning Technical Report
https://arxiv.org/abs/2412.01253
It was already discussed 2 months ago:
Eric Schmidt DROPS BOMBSHELL: China DOMINATES AI!
https://www.youtube.com/watch?v=ddWuEUjo4u4
Bye
Throwing papers at us (i.e. in usenet) although better than nothing,
it still is an extremely inefficient way to communicate.
Speak, yourself, in the way that a first year undergraduate prof
explains stuff to the students who just a few months earlier finished
their high schooling.
Then you will communicate.
If it is too difficult to do that (cause requires deep understanding
of the subject by yourself), or, if you don't get time doing all
that, then break the job down into little pieces. A day at a time, or
a week at a time.
You'd be spending the same amount of time as you are now, and yet you
will actually communicate something with those with physics background.
Did I have to say this? I don't think so. But your intentions here
are vague. But if you need support for your own gains in
understanding the subject, you just have to do it the way I explained
above. Then you will get that support. Not from Bozos and
"engineers", but from me and anybody with actual physics background
and such trainings.
There are a few actual physicists in relativity group that can join
in, but only if you do it the way I explained above.
- Hindu
- Jewish
- Liberal
- Asian
Hi,
The biggest group morons are those that
dream that they can profit by relocating
from their own group of people,
to a group of american people, and then
cry when their assets get frozen or who
knows what happens. The papers I posted
are about automatically tuning into groups:
Group Preference Optimization
https://arxiv.org/abs/2310.11523
DeepSeekMath: Pushing the Limits
https://arxiv.org/abs/2402.03300
Its actually quite amazing how machine
learning can utilize group identification
to customize how a conversational software
agent interacts. Currently I think is a
variant of fine tuning, but on a different
model level that is called policy. The first
paper has a figure with the following groups:
- Hindu
- Jewish
- Liberal
- Asian
Have Fun!
Bye
Mild Shock schrieb:
Hi,
Typo: now -> new
to generate now thinking
Should read:
to generate new thinking
I shit and piss on some asslicking group
that cannot think on his own. Thank god
USENET has only rarely such group morons.
Bye
Mild Shock schrieb:
Hi,
You wouldn't notic that I speak for myself
always and ever. Just remove the links and
you see its all text original by me.
I always post geniun commentary to current
events which are solely the product of my
own mind. I do not intend to educate something.
My intention is to document my current
understanding and to generate now thinking.
If you have expected some schooling, I am very
sorry for you. I never post boring schooling.
Bye
---------------- cut here -----------------
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
---------------- cut here -----------------
Physfitfreak schrieb:
On 2/4/25 3:06 AM, Mild Shock wrote:
Hi,
Because of the wide availability of Machine Learning
via Python libraries , the whole world (at least China)
has become a big Petri Dish that is experimenting with
new strategies to evolve brains on the computer.
Recent discovery seems to be Group Preference Optimization.
This is when you make the chat bot, detect and react
differently to different groups of people. It seems to
work on the "policy level". I don't understand it yet
completely. But chat bots can then evolve and use
multiple policies automatically:
Group Preference Optimization
https://arxiv.org/abs/2310.11523
DeepSeekMath: Pushing the Limits
https://arxiv.org/abs/2402.03300
Now it seems that it is also at the core of DeepSeekMath,
what is possibly detected is not group of people, but
mathematical topics, so that in the end it excells.
When unsupervised learning is used groups or math
topics might be found from data, through a form of
abduction.
Bye
Mild Shock schrieb:
Hi,
Wait till USA figures out there is a second
competitor besides DeepSeek, its called Yi-Lightning:
Yi-Lightning Technical Report
https://arxiv.org/abs/2412.01253
It was already discussed 2 months ago:
Eric Schmidt DROPS BOMBSHELL: China DOMINATES AI!
https://www.youtube.com/watch?v=ddWuEUjo4u4
Bye
Throwing papers at us (i.e. in usenet) although better than nothing,
it still is an extremely inefficient way to communicate.
Speak, yourself, in the way that a first year undergraduate prof
explains stuff to the students who just a few months earlier
finished their high schooling.
Then you will communicate.
If it is too difficult to do that (cause requires deep understanding
of the subject by yourself), or, if you don't get time doing all
that, then break the job down into little pieces. A day at a time,
or a week at a time.
You'd be spending the same amount of time as you are now, and yet
you will actually communicate something with those with physics
background.
Did I have to say this? I don't think so. But your intentions here
are vague. But if you need support for your own gains in
understanding the subject, you just have to do it the way I
explained above. Then you will get that support. Not from Bozos and
"engineers", but from me and anybody with actual physics background
and such trainings.
There are a few actual physicists in relativity group that can join
in, but only if you do it the way I explained above.
| Sysop: | Keyop |
|---|---|
| Location: | Huddersfield, West Yorkshire, UK |
| Users: | 468 |
| Nodes: | 16 (2 / 14) |
| Uptime: | 15:23:23 |
| Calls: | 9,440 |
| Calls today: | 3 |
| Files: | 13,591 |
| Messages: | 6,109,504 |