Why does my XOR ANN fail to generalize? [closed]

This is the octave/matlab code I have that calculates the XOR [0 0 -> 0; 0 1 -> 1; 1 0 -> 1; 1 1 -> 0]. The problem is that it fails to generalize/converge and calculates a small error for [0 0] & [1 1] and a really big one for the other 2. What is the issue with the code or the algorithms implemented? This is my first time making an AI and it feels that I am missing out on something quite obvious. Thanks,

BTW, the graph at the bottom shows the error with (Yexpected-Yresult). I know that there are other methods for managing the error/cost such as sum(error^2)/2 but I’ve decided to keep it simple. Network topology: 2 (input layer) -> 3 (middle layer) -> 1 (output).

clear  graphics_toolkit("gnuplot") sigmoid = @(z) 1./(1 + exp(-z)); sig_der = @(y) sigmoid(y).*(1-sigmoid(y));   function [cost, mid_layer, last_layer] = forward(w1,w2,b1,b2,data,sigmoid,i)   mid_layer = sigmoid(sum(data(1:2,i).*w1)'-b1);   last_layer = sigmoid(sum(mid_layer.*w2)-b2);   cost = data(3,i)-last_layer; end  function [w1, w2, b1, b2] = backprop(w1,w2,b1,b2,mid_layer,last_layer,data,cost,sig_der,sigmoid,i)   delta2 = (sig_der(last_layer,sigmoid)).*cost;   delta1 = (sig_der(mid_layer,sigmoid)).*sum(delta2.*w2);   w2 = w2 + 0.00001 .* mid_layer .* delta2;   w1 = w1 + 0.00001 .* data(1:2,i) .* delta1';   b1 = b1 + 0.00001 .* delta1;   b2 = b2 + 0.00001 .* delta2; end  data(:,1)=[0; 0; 0]; data(:,2)=[1; 0; 1]; data(:,3)=[0; 1; 1]; data(:,4)=[1; 1; 0];  w1=rand(2,3); w2=rand(3,1); b1=ones(3,1); b2=ones(1,1);  for j=1:20000   for i=1:4     [cost, mid_layer, last_layer] = forward(w1,w2,b1,b2,data,sigmoid,i);     [w1, w2, b1, b2] = backprop(w1,w2,b1,b2,mid_layer,last_layer,data,cost,sig_der,sigmoid,i);     cost_mem(j,i)=cost;   end end toc 

enter image description here enter image description here

The 2 graphs shows the error (sum(error^2)/2) for cases number 1 and 2.

Was there a published adventure that dooms the party if they fail a knowledge check?

I’ve recently seen a very strong claim made about knowledge skills:

“If someone didn’t take Knowledge: Nobility and Royalty and they can’t identify hearldry, then you can wind up getting ****ed. This can actually happen in a published adventure by the way.”

It’s possible that this was only referring to a minor issue like not being able to progress a conversation, but was there ever a published adventure that locks the party in to a bad ending if they fail a check on an uncommon Knowledge skill (e.g. Knowledge: Nobility and Royalty)?

How to let my players fail their rolls intentionally, but covertly?

I’m experimenting with some pen-and-paper role-playing involving characters with hidden agendas. To foster a proper sense of paranoia, I want these hidden agendas to be unknown to not only the other PCs, but the other players, too. A PC’s hidden agenda is known only to the GM and their player.

One of ways the secret agenda can manifest in the game is through the PCs intentionally failing tasks that would contradict their hidden agenda. For example, an undercover agent in a terrorist organization could try to aim their shots wide of the intended target, pretending to just have a bad day with their aim. Mistakes happen!

I want a rolling system that allows my players to intentionally make “unfortunate” mistakes without automatically implicating them of such treachery to the others. Here’s a list of criteria I’d like the system to fulfill:

  • As little overhead over normal rolling as possible: no pen and paper public key crypto!
  • The system must keep intentional failures indistinguishable from unintentional ones, at least between the non-GM players. Relying on suspicious gestures like note-passing or secret signals should preferably be avoided.
  • In particular, there mustn’t be a way for a player to prove that their roll was honest, lest the players decide to require such proof from everyone after each roll.
  • Preferably, the system makes it possible for a player to cheat downwards only – failure can be arranged, but success still requires luck with the dice!
  • Preferably the players roll their own dice – it’s not a huge deal, but getting to roll has a certain feel to it!

Does anyone have a system they have used or seen used that works given these criteria?

Note: Methods using cards and other randomization devices are also acceptable in lieu of dice, as long as the method proposed meets the standards outlined above.

Conjure (Minor) Elemental into a Hallow spell, does it just fail?

Hallow states “Celestials, Elementals, fey, Fiends, and Undead can’t enter the area…” but it doesn’t say what would happen if someone tried to summon some into the area. Would the spell fail? Would the elementals appear but then be forced to leave?

Would the ruling be the same for both Conjure Minor Elementals and Conjure Elemental? The latter uses a source of the element (e.g. a fire, what if the fire is in the Hallow area?) while the former just makes some minor elementals appear from an unspecified source.

Experiments Always Fail

I started using scrapebox 2 weeks ago. nothing I have done has worked except the keyword scraper.

I was trying to find an available domain through domainlookup. and an error message indicates a timed connection error. can you help what should i do?
Sleepy
https://prnt.sc/qh27vx
https://prnt.sc/qh284u

Why does Volatility fail on windows 10 dumps and what other tools can I use?

So I am trying to extract data from a full memory dump (Made with either dumpit or a BSOD). WinDBG manages to extract some information from it, but Volatility is silent:

 PS F:\> C:\Python27\python.exe C:\Python27\Scripts\vol.py -v -f .\DESKTOP-1NHUJ5K-20200115-133054.dmp imageinfo Volatility Foundation Volatility Framework 2.6.1 *** Failed to import volatility.plugins.registry.shutdown (ImportError: No module named Crypto.Hash) *** Failed to import volatility.plugins.getservicesids (ImportError: No module named Crypto.Hash) (More unrelated import warnings) *** Failed to import volatility.plugins.registry.shimcache (ImportError: No module named Crypto.Hash) INFO    : volatility.debug    : Determining profile based on KDBG search... WARNING : volatility.debug    : Alignment of WindowsCrashDumpSpace64 is too small, plugins will be extremely slow  WARNING : volatility.debug    : Alignment of WindowsCrashDumpSpace64 is too small, plugins will be extremely slow (repeated a dozen times) WARNING : volatility.debug    : Alignment of WindowsCrashDumpSpace64 is too small, plugins will be extremely slow           Suggested Profile(s) : No suggestion (Instantiated with Win10x64_15063)                      AS Layer1 : SkipDuplicatesAMD64PagedMemory (Kernel AS)                      AS Layer2 : WindowsCrashDumpSpace64 (Unnamed AS)                      AS Layer3 : FileAddressSpace (F:\DESKTOP-1NHUJ5K-20200115-133054.dmp)                       PAE type : No PAE                            DTB : 0x1aa002L              KUSER_SHARED_DATA : 0xfffff78000000000L            Image date and time : 2020-01-15 13:32:57 UTC+0000      Image local date and time : 2020-01-15 14:32:57 +0100  PS F:\> C:\Python27\python.exe C:\Python27\Scripts\vol.py -v -f .\DESKTOP-1NHUJ5K-20200115-133054.dmp --profile=Win10x64_17763 pstree Volatility Foundation Volatility Framework 2.6.1 *** Failed to import volatility.plugins.registry.shutdown (ImportError: No module named Crypto.Hash) *** Failed to import volatility.plugins.getservicesids (ImportError: No module named Crypto.Hash) (more irrelevant import errors) ... Name                                                  Pid   PPid   Thds   Hnds Time -------------------------------------------------- ------ ------ ------ ------ ---- 

Note that the pstree function prints no error and no garbled output: It just exits.

All the import errors are simply caused by the lack of the PyCrypto library but have no bearing on the functions I used.

I know for a fact I am using the correct profile since I made the dump myself on a VM running Windows 10 17763.

All of these seem to work perfectly on older versions of Windows. Is there a reason for Windows 10 to work differently with volatility?

My purpose is to extract a binary executable file from a full crash dump. Is there any tool other than volatility that can allow me to do that?

Experiments Always Fail

I started using scrapebox 2 weeks ago. nothing I have done has worked except the keyword scraper.

I was trying to find an available domain through domainlookup. and an error message indicates a timed connection error. can you help what should i do?
Sleepy
https://prnt.sc/qh27vx
https://prnt.sc/qh284u

How does progress fail in system $F_{\omega}$ when types $T_1 \to T_2$ and $T_2 \to T_1$ are equivalent?

Pierce’s TAPL book gives in exercise 30.3.17 the setting where $ T_1 \to T_2 \equiv T_2 \to T_1$ (the function type are assumed to be equivalent). In the solutions, he claims that this assumption breaks the progress property.

It is easy to see that preservation fails. How can progress:

$ \vdash t:T \implies t \text{ is a value } \lor \exists t’. t \to t’$

be wrong in this setting?

If you cast a higher-level spell from a spell scroll with a casting time of 1 action, but fail the check, can you cast a bonus-action spell that turn?

The bonus-action spellcasting rule states (PHB p. 202):

A spell cast with a bonus action is especially swift. You must use a bonus action on your turn to cast the spell, provided that you haven’t already taken a bonus action this turn. You can’t cast another spell during the same turn, except for a cantrip with a casting time of 1 action.

The description of a spell scroll states (DMG p. 200):

If the spell is on your class’s spell list but of a higher level than you can normally cast, you must make an ability check using your spellcasting ability to determine whether you cast it successfully. The DC equals 10 + spell’s level. On a failed check, the spell disappears from the scroll with no other effect.

If you fail the check to cast the higher-level spell via the scroll, does the spell count as being cast despite there being no effect?

This question is very similar to this one – If a spell cast from a Spell Scroll is Counterspelled, is the scroll consumed? – but I am not sure if the answers in that question apply to this situation as well.