Greetings Codeforces!
sammyuri, myst-6 and I are absolutely chuffed to bits to invite you all to have a go at Codeforces Round $$${12}_{10}$$$ (Spectral::Cup 2026 Round 2 (Codeforces Round 1100, Div. 1 + Div. 2)), kicking off at May/23/2026 17:35 (Moscow time).
There will be 9 absolute cinema problems for you to solve over the course of 3 hours, some problems will have multiple parts. This round will be combined for Division 1 and Division 2 and will be rated for everyone.
I’d like to raise a glass to the following brilliant minds for making this whole thing possible:
- Dominater069 for splendid coordination.
- Um_nik for preliminary review.
- The fine assembly of testers: arvindf232, EvenImage, Geothermal, Sana, Wansur, p0tato, __baozii__, Aotsuki, Dragos, catgirl, Monogon, nifeshe, Proof_by_QED, lucaski2, efishel, is_i, Argentum47, reirugan, naneosmic, chromate00, Nyemot, simplelife, eric899.
- moonpole for remembering to test.
- Alexdat2000 for Russian translation.
- MikeMirzayanov and KAN for Codeforces and Polygon.
- You for participating.
Score distribution: $$$500 - 750 - (750 + 1250) - 2000 - 2500 - 2500 - 3250 - 4000 - 5000$$$
UPD: editorial
UPD2: Congratulations to the winners!
Now a few words from our sponsor.
Spectral::Technologies is an HFT fund – we build trading strategies and low-latency infrastructure for global markets. The people doing this: IMO, IOI, IPhO and All-Russian Olympiad medalists and top engineers – people who love the challenge and always want a bigger one. That's who Spectral was built for.
We are actively hiring! Check out our Quant roles:
- Junior Quant – 150–180k USD + profit sharing
- Middle Quant – 240–320k USD + profit sharing
- Senior Quant – 360–500k USD + profit sharing
We’re also hiring for C++, ML, and DevOps roles. Complete the application form to explore career opportunities with us.
We are excited to sponsor our second Round as part of Spectral::Cup 2026 — a three-round tournament where prizes get more valuable with every round.
Prizes for Round 2:
Top 30 will get laptop cases (390x290mm) + Codeforces-inspired anti-stress toys.
Make sure that you take part in at least two rounds to improve your chances to win the main prizes. We prepared bigger prizes for top performers by their final score in Spectral::Cup 2026. The final score is the sum of your best 2 results according to the GP500.
- Top-3 by final score will get (1st) MacBook Pro, (2nd) iPad Pro, (3rd) Whoop — or USDT equivalent to the prize value
- Top-15 by final score will get Claude subscription
- 30 random participants among top 500 based on the final score will get additional prizes
Code fast, think faster – see you in the next round of Spectral::Cup 2026!
As an author, I can confirm this contest is absolute cinema.
I confirm.
u gay
Кто?
As a participant, hope to become GM this time.
As a participant, hope to become expert this time.
orz
as a participant i hope i hope
Failed :(
As a tester I must keep creating memes:
Looking at the score distribution, and bearing in mind the standard of the last match, I reckon this one is going to be absolutely brilliant!
I hope I can hold onto my Master status.
As an author.
As a te
As
ㅤㅤ
\b
When did I test this round?
Maybe when yiren was using wavelet matrix
Nife's divorced mother just made another div
absolute cinema
As nifeshe's stepmother (Intellegent's current wife), I tested this round.
Top 30 will get an ellipsis... w problem setters
also it says codeforces round 12 lol
my first rated contest I'm excited I can do it but it conflicts with leetcode biweekly (w ragebait?)
It's round $$$12$$$$$$10$$$ = $$$1100$$$$$$2$$$
my rating will be absolute cinema after this contest :)
Wish to solve problem E
Mysterious Retired Man, Kuro_neko orzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz!
I'm a fan of Kuro_neko awa
I'm a fan of MaxBlazeResFire qwq
I know you on leetcode
What a coincidence! Could you please tell me your leetcode username?
I don’t recognize your Codeforces handle lol.
Kuro_neko will ak Div1+2
what's wrong with you...
As
no interactive problem?
yea it suckssssssss (I got no clue how to solve interactives I'm just farming contribution lol)
But I like them!
I am heartbroken... no interactive problem (not that I can solve them)
at least I get +Delta regardless of how many I solve
as someone who adores constructive algorithms I think I'll gain +100
I'M JUST WAITING.
absolute cinema lol im jk
The final score is the sum of your best 2 results according to the GP100 scores.30 random participants among top 500 based on the final score will get additional prizesSo if there are 3 contests, and only the top 100 in each contest get points according to GP100, assuming there are no repetitions, there can be at most 300 with positive scores. So how will there even be a top 500?
As someone guaranteed to tie for a rank better than or equal to 301st, I really hope I win a prize!
Hi Papaulo,
Thank you for being super observant. We will use GP500 instead. It has now been reflected in the announcement.
As a participant, I hope to become a specialist in this contest.
As a programmer, I hope you got lesson and will never cheat in upcoming contests.
will there be more minecraft problems? :3
As a as,as as As.
Round 1100! Incredible milestone for the community <3
As a participant , I can't confirm this is Absolute Cinema , but hope that it will be an Absolute Cinema!!
Whoever gets the iPad Pro, gift me pls :uwu: :smile:. You can dm me anytime :lol:
Maybe try this?
No money sirs
Maybe try this?
Haww! Police come and captivate, wrong preacher! I don't do bad if I can't do any good.
PS — Joking
Everyone, have a nice day༼ つ ◕_◕ ༽つ
Good luck to all participants!
death game
As a newbie, I can confirm my rating graph will become absolute cinema after solving problem A.
hope lowRatingProgrammer remain lowRated after this absolute cinema contest.
bro your official and max rating is 1299, but current rating is showing 1410? Can you tell me your secret of getting rating?
He got rolled back
bro is cheater. How I didn't notice it from the beginning. I thought you were above of this habit.
As a participant, I hope to become absolute cinema.
After a long busy time, I hope I can attend this.
hope to have a good cf round and hope to save expert this time (LOL)
This is the last Legendary Binary round that we can do ;(
only nigg@ join this contest
Hoping to solve A and B
Hope to gain rating on my 69th rated contest!! ^_^
nice
who can wish me that I will solve the problem B!
As a pupil, I hope I don't panic and mess up this contest TwT
bruh, its perform the operation many times (possibly none). But still funny tho :)
I fucking suck dude.
Problems were super fun, congrats to the setters!
Who else got tricked into thinking $$$C2$$$ is a DP?
So it is not DP??
Actually, I think C2 can be solved by DP, see my solution. I don't think it's the nicest solution ever but it passed the pretests.
Edit: link https://codeforces.me/contest/2229/submission/375829590
Prefix sum can be used to solve this problem so it's not totally wrong
This is the highest point I've ever had in a Div.2. I don't know if this was intended, but I think that this contest is easier than normal.
Problems were great but to me C2 > D. And how to solve E?
how u solve d
Binary search on the answer. Take a look at my AC submission
Hint : Root the tree at the node "n". For a node "x" to be in the set it all of its subtree has to be deleted by help of some other node "y" whose value has to be greater than or equal to values of all nodes in "x"s subtree (so that "y" can delete them).
really C2 was easy for me how do you get insights in D's in general i had 1hr20mins still couldnt do D what rating do you practice
To me when we have the answer always in range [1, 2n], the idea of binary search on answer is quiet obvious. Everything else I had to do was to check if we can construct a way to have a value X as answer or not.
binary .. how do you practice like daily
Was feeling bad about my performance until I realized Div1 and Div2 were combined
Can anyone please tell me that whether prefix suffix construction was only solution for C2 ? Most of the solutions at top tried this approach but mine was different... I was trying constructing a maxcount mincount logic where I was using 4 possible outcomes tracking and getting to answer by proper construction of max and min , then from last index check for flip or not based upon extra array storing ( flip means assigning -min to max and vice versa ) ..
My approach: Find the optimal +ve that I can make -ve while keeping all its prefix +ve. Try to make all its left elment negative so that in the last operation I can flip.
To keep all negative prefix flip from the last idx of positive segment then flip the last idx of the previous negative segmant (that already became +ve).
How to optimize F from $$$O(3^n \log a)$$$ to something like $$$O(n 2^n \log a)$$$?
my sol is $$$O(3^n+2^n*n)$$$ so it's a completely different dependence from what you're thinking about :D
By an exchange argument you can assume the last element used brings the min to being the max. The outer loops are iterating over the last element and doing binary search to find the highest reachable min from the remaining $$$n-1$$$ elements. Inner loop is a bitmask DP that maximizes (buckets filled, size of current bucket) lexicographically. Note this is actually $$$O(2^n n^2 log(a))$$$
Seems to work, thanks!
I liked all the problems I attempted (A B C1 C2 D E F). I wish I was able to solve all of them.
I liked the transitioning between C1 and C2 but it was hard for me. I think C2 score could be a bit higher.
Thank you for the short and clean statements. Great contest!
Loved the problems, especially D. Thanks to authors!
how to become good at reading cf problem statements? I solved the pretests at C1 oblivious to the condition that you can only select an $$$i$$$ if
a[i] > 0, so when I tried to solve C2... got WA on pretest 1, I checked the example inputs and outputs and wasted time being frustrated for maybe ~30 mins on why the output is not making any sense, only after re-reading the problem a dozen time I noticed the condition but then I only have like ~3 mins left to think about the solution.Without that constraint, C1 and C2 would use exactly the same solution.
So I think the general answer is: make sure to read both parts of the problem and try to understand how they're different, even before solving part 1.
After this contest, i can say this is absolute cinema
A rather similar problem to I is https://qoj.ac/problem/4815
The solutions are almost the same
Which problems had pretests!=systests?
No idea. It seems there were not any hacks, so systests = pretests for every problem?
I'm asking about initial setting, excluding hacks. I've resubmitted the code (and that costed me a place) being afraid of systests being stronger than pretests and seeing no annoucement about pretests = systests, while it seems that it was pointless again. I think I've touched the topic of pretests vs systests during the first round of spectral cup, maybe it'd be good to comment on it somehow?
For example say that from now on always systests = pretests + unexpected hacks (if that was the case) and that the solution is simply judged twice (during and after the competition) and it needs to pass both times to get AC? I think people quietly assume the former and not assuming that (because I think it isn't written anywhere, correct me if I'm wrong) is a bit painful, while I still consider it the only valid approach. About the latter, it's mostly about rejudging solutions manually (because it seems that it has happened last time), as doing so is wrong.
Obviously, if mostly systests = pretests it would be the best to somehow improve the judging system (to not judge twice), but I'm aware that that's additional work, so I'm looking for an easier solution.
Excuse. I solve problem H, but the submission 375835599 get skipped, same for 375836751. I can guarantee that I solve this problem with no cheating, and I do not receive any message like "Your code is similar to another submission". How to appeal? I don't find any help about this. If need, I can provide the coding history in VSCode.
If you can give me some suggestions, thx.
obfuscation of code is not allowed
No obfuscation, you can view my other codes, I use this coding style, though I know it is uncommon.
It doesn't become obfuscation any less just because you do it all the time. Your code should be at least somewhat readable to others (and yourself) because reading others' code is (at least in theory) part of CF contests. It doesn't matter if you code that way by hand even though it sounds like bullshit that anyone would ever manage it; run your submits through a formatter or something if you don't want to have trouble with that rule.
Yes. But if I just need to format the code, why hacker not? It is hacker's responsibility to understand, but for hacking (even just reading), it is not friendly to prepare a 20KB sample code, as any hacker doesn't want to judge the correctness of the sample code.
Obfuscation is not allowed, but if anyone can read it by formatter, I think it is no reason to skipped a submission because participant doesn't format the code.
When you hack other's solutions, you do not have access to the code in the text form; you are given the code as a picture. You cannot run formatter on other's code, and you basically need to judge other's code with your eyes only. (That's why we have "no obfuscation" rule.)
Emm, ok, but I don't think if a code is hard to hack, it should be skipped. Thanks for your explanation.
Isn't putting several hundred lines of code at the beginning of each solution an obfuscation as well?
I think that your submission should be unbanned, maybe it is worth to tag MikeMirzayanov, Vladosiya, KAN and also write someone idk
But anyway, consider changing of codestyle, maybe with better codestyle you will make less bugs)
Very thx for your tag! I don't know who I should tag for appeal before.
I get used to and enjoy this codestyle in 4 to 5 years, but I may try to change it after, for more contests with ACM-ICPC rule. None of ACM participant wants a teammate using this codestyle :)
This was my first contest on codeforce. I hope to learn more with consistent participation and practice and meet great people.
Good luck!
I believe some bug happened during system testing?
My solution for F initially passed the pretests but got compiler error during system test. Can you look into this problem?
375848558
I rejudged it and it became TLE. I think it would be unfair to rejudge it again since others can't rejudge their TLEs on system tests. (I also submitted it a few times myself and it got TLE consistently.)
There's definitely been cases of rejudging TLEs on system tests (in particular, when the TLE is on a pretest that the code previously passed). Does CF have a well defined rule for how many rejudges one is given in such cases?
Interesting, I didn't know about this case. About a year ago I remember admins skipping a submission after the authors rejudged it because of TLE on system tests, so I assumed it was the standard to not rejudge such cases. If something has changed I would be happy to rejudge the submission again.
Genius level problem-solving!
https://codeforces.me/submissions/darshan772003
Aspire to be like this guy one day (or not)
I was checking random standing pages of people who got 1200th place or worse. It's all green and cyan. That's crazy how cheating ruined the sport for lower rated participants.
Man i just want to reach pupil.
Good contest
I think there are many cheaters in this contest, and I hope CF can address this issue.
hi
This problem was pretty easy and tricky...and really this problem is absolute cinema..
Is there any wrong in test cases of C2?? thanks
can i know test and why this wrong ?
You haven't even submitted to C2? I think the test cases are fine
System MikeMirzayanov I received a coincidence warning for my submission 375835571 on problem 2229F, with submissions 375817492 and 375828452.
I would like to clarify that I do not know these two users and I did not share my code with them, nor did I receive or copy code from them during the contest. I also did not use any public online IDE or paste service that could have exposed my code.
My solution was written independently. The idea and implementation for this problem are relatively simple and standard, so I believe similar code structure may naturally appear among independent submissions. In particular, the main logic does not rely on any unusual trick or custom implementation that would indicate copying.
I understand that unintentional leakage is also considered a violation, so I am providing this clarification here. If needed, I am willing to provide any additional details about my solution process.
Obviously you used AI according to strange changed and AI-ish coding styles and AI generate similar codes for the same problem. Also for many submissions of yours, ChatGPT generates almost same solutions. Another IGM cheater?
Additionally your comment is also AI-generated. Scanned by GPTZero.
I suggest trying screen recording for the whole process and closing QQ. I was skipped before because I didn’t close QQ, which led to my code being leaked.
Why is everyone downvoting? I don’t understand. I was just making a very reasonable suggestion that could help reduce cheating and increase the chance of a successful appeal.
Screen recording is not enough — you can ask AI by using your other devices, and only rewrite (copy) the code on your recorded device. That's why AGC proctoring rules (https://atcoder.jp/posts/agc_proctoring_en) requires dual-camera recording (entire screen & yourself).
Closing QQ is nonsense. Messaging apps will never proactively copy your code and transmit to others. Only doing broadcasting on QQ, or you sending code to others, can leak your code.
First of all, I think doing these things is at least better than doing nothing. At least, it would help avoid situations like XVIII, where a single problem receives dozens of submissions that fail on Test 1.
Also, closing chat apps is meant to reduce suspicion of cheating and to make sure, as much as possible, that there are no conversations about the problem.
Hello MikeMirzayanov I received the coincidence warning regarding my solution 375795858 for problem 2229D.
I would like to clarify that I solved the problem independently during the contest and did not intentionally copy code from any participant or external source. I do not know the mentioned users personally, and I neither shared my solution with anyone nor received code from others.
I also did not use any public IDE, paste service, or platform that could unintentionally expose my code. The approach for this problem is fairly standard, and I believe similar implementations can naturally occur between independent submissions, especially in competitive programming contests where many participants arrive at the same logic.
I fully understand that even unintentional leakage is considered a violation, so I wanted to provide this clarification immediately. If needed, I am willing to explain my complete thought process and implementation for the solution.
No matter this is a coincidence or not, anyway you used AI. And everyone knows the same model often generate similar codes for the same problem.
https://codeforces.me/contest/2229/submission/375824224 & https://codeforces.me/contest/2229/submission/375827759 -> One use lambda DFS, one use traditional function DFS. WA on pretest 1 (sample). No one will submit two sample-incorrect solutions in a row, without any local testing.
https://codeforces.me/contest/2229/submission/375815875 -> This is the most AI-ish. long, detailed var names differ from your other codes;
if (cin >> t)AI-favored behavior.the worst contest ever
Hello. The submissions flagged for similarity are:
Both accounts belong to me. I accidentally submitted the same solution from different logged-in browsers/accounts during the contest. There was no collaboration or code sharing with any other participant.
I understand this violates the rules and it was my mistake.
Hello MikeMirzayanov
Regarding the plagiarism flag for my submission (https://codeforces.me/contest/2229/submission/375847735) on problem 2229D:
I would like to clarify that my solution was derived independently. This problem is a classic "Maximize the Minimum" task, which is generally a template for Binary Search on Answer type problems. The same template can be found on questions with Maximize the Minimum or Minimize the Maximum type problems as listed below : Split Array Largest Sum. This template requires finding a candidate value and then simulating if the given scenario is possible with the candidate value.
Next comes the checking algorithm I used, the idea was to see each value as 0,1,2 based on whether they are lesser than candidate value, or one is less one is greater or both are greater than or equal to. The idea of this new array or 0,1,2 originated because in the previous question i.e 2229C-MySolution I used a target array S based on whether the current sign of our original array is positive or negative and hence an array of 0 and 1s. Next my D array which is basically the the pair wise XOR of the problem. Next how to we check whether candidate is possible or not, it was evident that value 1 doesn't make any change to our answer, only 0 and 2 make a difference. Thus we wanted to have number of 2's more than the number of "zero_blocks". The thought process behind this also originate from the solution of the problem 2178A where a N in the array doesn't affect the number of Y.
I don't know how my question has been flagged for plagiarism. The idea and logic behind my solution is a mixture of all concepts used in the above mentioned questions.
I request the concerned authorities to look into this matter. Regards pekka_packer
I am writing regarding the plagiarism warning I received for submission 375832332 on problem 2229D.
I want to clarify the following points:
The warning states that my submission coincides with rustagimohit48/375832332. However, rustagimohit48 is MY OWN account. The same submission ID (375832332) appears under both accounts, which appears to be a system error in the plagiarism checker.
Regarding ByteMystic — I do not know this person at all. I have never communicated with them, shared code with them, or seen their solution.
Problem 2229D involves binary search, which is a very standard algorithmic technique. Binary search solutions for the same problem naturally tend to look extremely similar across different participants — same logic, same structure, same boundary conditions. This is a well-known issue with plagiarism detection on implementation-heavy problems.
I did not use any public code sharing sites (Ideone, Pastebin, GitHub, etc.) during the contest. I wrote my solution entirely on my own.
I kindly request the admins to review this case, especially the fact that the plagiarism checker flagged my own submission ID under a different username, which suggests a detection error. I have been genuinely trying to improve my competitive programming skills and do not want my progress affected by a false flag.
Thank you for your time and consideration.
.
Sorry for copying solutions during the contest. I understand this breaks the rules. It won’t happen again. Please forgive me and remove the warning/block if possible.
I promise that this will not happen again.
Where can I check my rating?
There are standings on the competition page
Codeforces inspired antistress toys :d
Absollute Codeforces
Hello MikeMirzayanov and Codeforces administrators,
I would like to respectfully clarify my situation regarding the plagiarism warning for my submission:
https://codeforces.me/contest/2229/submission/375839389
The warning mentions similarity with:
https://codeforces.me/contest/2229/submission/375827337
and
https://codeforces.me/contest/2229/submission/375826992
I want to clearly state that I wrote my solution independently during the contest. I do not know these users personally, I never communicated with them during or before the contest, and I never shared or received code from anyone.
After carefully comparing the codes, there are multiple implementation differences between my solution and the compared submissions.
Some important differences are:
My implementation uses arrays
reqandM, while the other submission usesDandMaxBwith different DFS transitions.In my DFS:
while the other submission computes:
These are structurally different state definitions.
while the other submission explicitly marks nodes using an additional DFS:
This is a major logical and implementation difference.
Variable naming, subtree handling, DP transitions, prefix computation, and answer construction are written differently.
My submission does not contain copied templates, macros, comments, formatting style, or identical structure usually associated with direct copying.
This problem naturally leads to a constrained tree-DP approach, so similar high-level ideas between independent contestants are expected.
I also want to mention that one of the referenced accounts appears to already be disabled, but I had absolutely no connection or interaction with that user.
I respectfully request a manual review of my submissions and contest activity. I genuinely participated fairly, and I would be very grateful if the warning could be reconsidered and my skipped submissions/results restored if possible.
Thank you for your time and understanding.
Hello . My account nikuu8191 appears to be disabled after a contest plagiarism warning. The duplicate submissions came from two accounts owned by me. I accidentally submitted the same solution from different logged-in browsers/accounts during the contest. There was no collaboration or code sharing with another person. I understand this still violates the rules, and I am sorry for the mistake. I would appreciate clarification regarding the status of my account. Handles: * nikuu8191 *24je0142 Submission IDs: * 375794833 * 375793294
BEI BEI XVIII
Hi, I wonder how the prizes can be claimed? Thanks.
Hi, Top 30 in Round 2 will be announced shortly. We will then contact them for delivery details and pass the information to Spectral Team. Once the prizes are shipped the recipients will get tracking details. Please bear in mind that production of branded items can take up to 8 weeks. Thank you.
Thank you!
Congratulations to the winners of Spectral::Cup 2026 Round 2
You will be contacted via private messages with instructions to receive your prize.
Top-30:
Wow!!!!
BYE BYE XVIII
[Deleted,somebody stole my account]
Dear coordinators of Round #2229
My name is Shivam Chahal (primary handle: Chahal25 ). I am writing to clarify a mistake I made during the recent contest: Spectral::Cup 2026 Round 2 (Codeforces Round 1100, Div. 1 + Div. 2).
Both accounts Chahal25 and Shivam_Chahal belong to me. I created and used two profiles unintentionally, and I submitted solutions from both accounts. I solved Problems A and B myself, ie 2229A - Slimes on a Line and 2229B - Absolute Cinema ; the code on both accounts is my own work. There was no code sharing with other participants, no plagiarism from external sources, and no intent to gain an unfair advantage.
Link to my submission on primary account : 375798916 and Link to my submission on second account : 375798534
I understand that using multiple accounts during a contest violates the rules. I sincerely apologize for this lapse in judgment and accept responsibility. To comply with the rules, I am permanently abandoning the Shivam_Chahal account and will use only my primary account (Chahal25) for all future contests.
Because the solutions were originally authored by me, I respectfully request that no further action be taken against my main account beyond the warning already issued.I will accept and respect whatever decision the administration deems appropriate.
Thank you for your time and understanding.
Regards, Shivam Chahal (Chahal25)
Following is the email i received from the codeforces: Attention! Your solution 375798534 for the problem 2229B significantly coincides with solutions Shivam_Chahal/375798534, Chahal25/375798916. Such a coincidence is a clear rules violation. Note that unintentional leakage is also a violation. For example, do not use ideone.com with the default settings (public access to your code). If you have conclusive evidence that a coincidence has occurred due to the use of a common source published before the competition, write a comment to post about the round with all the details. More information can be found at http://codeforces.me/blog/entry/8790. Such violation of the rules may be the reason for blocking your account or other penalties. In case of repeated violations, your account may be blocked.