CSD/LCD Experiences - Medium Articles

How I Got Into IIIT Hyderabad (Anushka Pratap Singh, 2025)

Background: Heard about LEEE from a friend; was unsure at first (low acceptance rate; had spent half the semester preparing for Amazon OA, which she cracked). Decided to give it a shot.

Interview Strategy

Result on May 12; interview on May 24. Everyone whose interview transcripts Anushka read chose Maths as their “favorite subject.” She didn’t, because she wasn’t confident in it. Instead she brushed up on OS and DBMS - maybe 2-3 hours of prep for 3 days. Also tried prepping for NLP questions but wasn’t confident, so avoided making deep claims.

Key counter-strategy: Deliberately chose subjects most other candidates avoided rather than Maths, betting on genuine strength.

Interview Transcript (Verbatim)

Interviewer: Okay tell us about yourself and why do you wanna join IIITH?

Anushka: I’m Anushka, currently in IIIT Kottayam, not for more hopefully. I have done a few gigs in React. I will start interning in Amazon from the next week. About IIITH, it’s one of the best colleges of India with one of the best alumni networks. I currently am not knowledgeable enough in research work, but I would like to change that and contribute to research that matters. Also actually, last sem PK sir had visited our college to take a session.

(They interrupted.)

Interviewer: What did he talk about?

Anushka: He talked about how models are biased - like, if you ask for a photo of a nurse they would send a woman, but if you ask for a doctor they would send a man. And he was explaining why it happens, why it shouldn’t happen, and how we should go through these checks with our models and data. That was a really cool session.

Interviewer: Did someone answer the questions? Who told you about IIITH?

Anushka: A senior answered all his questions, also got an internship at IIITH through that, and he only told me about lateral entry.

Interviewer: Anushka, do you think you’re biased?

Anushka: Huh?

Interviewer: Do you think you’re biased towards IIITH or PK sir, since you said you only want to come here because of his session - so do you want to come here only because of one session?

(Anushka laughed at this.)

Anushka: Yeah I’m definitely biased towards IIITH, but not because I have only been fed positive data about it but because I’ve done my own research and I independently decided that this is the correct choice for me.

“I nailed it - they were charmed btw. Now that I look back, I realise how smart their question was.”

Interviewer: Okay okay, what are your research interests?

Anushka: I haven’t had the chance to explore much. I’ve studied the basics of NLP, watched and understood sessions of DeepLearningAI, tried reading research papers. I would like to explore more before deciding on what exactly I would like to work on.

Interviewer: Okay, let’s start asking questions then. What are your 2 favorite subjects?

Anushka: OS and DBMS.

(OS questions: race condition, difference between process and thread, context switching, deadlocks - she answered all. DBMS: what’s a table/tuple/row; difference between primary and unique key.)

Interviewer: Give an example of a successful and failed transaction, and why successful transactions are important.

Anushka: For example, imagine your salary is credited, but it doesn’t reflect in your bank account - that would clearly be a failed transaction. This is exactly the kind of scenario a successful transaction should prevent.

(They laughed - they liked the example.)

Interviewer: What are the ACID properties? Can you give real-life examples?

(Anushka gave examples but forgot the meaning of Consistency - due to nervousness. The interviewer tried to guide her. After the first hint, she could answer.)

Interviewer: Okay, that’s enough. Thank you Anushka.

Anushka: Huh? So fast? Okay thankyou.

“I didn’t know if I would get in. I answered like 8/10 questions, but from past interviews I’ve seen, even if you don’t answer all of them, people got in - so I had hopes.”

Final Thought: “If you’re reading this and wondering if you should take the risk - just do it. Worst case, you’ll learn a lot. Best case, you’ll be writing your own IIITH blog someday too.”

IIIT-H Lateral Entry Interview (Anmol Mishra, 2018)

Stats (2018): Around 42 students were called; only 37 appeared for CSD (remaining were absent).

Setup: One person called at a time. 5 professors sitting in a room. They asked Anmol for documents and asked if he had any subject preference - most students were not asked this. He replied: Data Structures and Algorithms.

Interview Transcript (Near-Verbatim)

DSA - Insertion Sort:

They asked what sorting is, then asked Anmol to explain insertion sort. He explained using an example of a pack of cards. They asked for the code. He tried, but they said it had a bug.

Anmol: I clearly stated that it may take a bit for me to figure out. I usually take code from Stack Overflow.

“They smiled and giggled among themselves” and then told him “OK, No issues.”

COA - Weakest Subject:

They shifted to Computer Organisation and Architecture.

Anmol: Sir, it is evident from my marks. I got C+ grade and it’s hard for me to recall.

A professor replied that he had the same grade in his undergrad and now has a Ph.D in COA. (“Maybe trying to pump me up.”)

DFA:

They asked a DFA question. Anmol tried halfway. They said: “You may have studied a bit long ago so you are near the answer but taking long to reach to answer, leave it. We have your thought process now.”

Mathematics (triggered by “I love Maths” in SOP):

Complex numbers: find root for $x^3=1$. Anmol answered. They asked him to “enlighten the process” and he stated it step by step. Then: “Can you find root $x^7=1$?” - he said YES. They stopped him and asked:

Interviewer: If you have $4.000000000000000000000000000000000001$, how will you store it on your machine?

Anmol tried bit manipulation. They told him to try something different. He kept thinking. They denied a custom data type.

Anmol: Sir, probably your vision is not approachable by me at the moment, it’s not hitting me, sorry.

They looked chill and said “We will answer this later” and laughed among themselves.

3×3 Matrix (binary values):

They asked him to write a 3×3 matrix with only binary values and determine if it was invertible. He said: to check invertibility, I need to find the determinant. They asked him to find it.

Anmol had not touched matrices in a long time; he missed the fact that when two rows or columns are the same, the determinant is zero. Instead he used his class-12 method to find the determinant.

“They were shocked at how I was finding the matrix determinant.” They asked him to explain his process.

Anmol shared how he was doing it. They looked fascinated but told him to think of a better approach. Then it hit him - he told them the fact about identical rows giving determinant zero. They said: “Great, this was appropriate.”

ML Section (triggered by “ML Enthusiast” on resume):

They asked him to go to the board: supervised vs. unsupervised learning, decision tree, decision boundary.

As soon as they shifted to regression:

Anmol: Sorry sir, in my summer training at IIT-Kanpur, I have not yet started with the regression model.

One prof: “Being an enthusiast, answer this - how will you store a speech signal in your system?”

Anmol: I am giving a try and I will use a supervised learning with wavelength for some time period like 1-2 ns and then label the incoming speech.

They did not seem convinced but praised him for the try.

The Continuous/Discontinuous Moment:

Interviewer: Is my voice a continuous or a discontinuous function with time?

Anmol: Sir, you are pausing in between words - what’s the consideration for those?

All professors just laughed for 5-6 seconds and their laughter was echoing. “I felt comfortable.”

Another prof said: “Let’s say a human is not speaking and a whistle is blowing.” - Anmol answered: “CONTINUOUS SIGNAL.” They asked him to go to the board and draw the graph. As soon as he sketched it in the negative Y direction they stopped him. He defended: “Whistling is going to produce a sound wave which is nothing but a pressure wave in form of compressed air particles, so pressure can be on the system and pressure can be away from the system.” They were convinced.

Signal processing: They then asked about Sampling, Aliasing, etc.

Digital Logic: They asked if he was comfortable with it. He confidently said Yes. They did not ask about it. They asked him to leave.

Key Observations

“Most students were interviewed for 15-20 minutes only. I was among the last 5; they had their lunch just before my interview and took mine for around 35 minutes.”

“They were looking forward to know how you think only and were not interested in how well you remember the formulae or topics - but will you be able to connect some random dots?”

Takeaway: Everything you mention in your SOP or introduction becomes fair game. But equally, honesty about gaps (“C+ grade”, “haven’t started regression”) is received well - professors respond with encouragement, not judgment.

My Lateral Entry Journey - IIIT Hyderabad (Aryaman Kolhe, 2022)

Tldr; Check out the LEEE official homepage for your year’s syllabus. Make sure your fundamentals are set and solve as many GATE questions as possible. For the interview, mention your current interests and what you have been up to in the last 2 years.

Preparation (Round 1 - Written)

• Primary resource: Acing the GATE, CS and IT (2019, Wiley) - found in his college library; used as a questions repository despite typos.
• Practiced GATE CS questions extensively (Geeks for Geeks, Gateoverflow).
• K.H. Rosen Discrete Mathematics and Its Applications for Discrete Math fundamentals (even Chapter 1 alone is very helpful).
• Strategy: focused on DSA and Discrete Math; spent less time on Digital Logic (worked for him but he explicitly warns this is risky).

Critical Warning: Digital Logic Risk (Aryaman Kolhe)

“I should mention here that although this worked for me (I got lucky), it might not work for many people. Sometimes the weights for different sections can change, and this year (2023), many people were asked questions from topics they did not prepare for in the interview. 2 very talented individuals that my friends and I knew did not make it because their interview panel stuck to questions in Digital Logic Design, which is something they did not prepare for. Try your best to cover all topics, but optimize on the ones you know best.”

• Solved Leetcode problems in C to practice DSA and pointer/logic thinking.
• 2023 additions: Mathematics section added (Rolle’s Theorem, Lagrange’s MVT, vector spaces and rank in linear algebra, graph theory - all GATE-level).
• Key principle: “Pay attention in your current university. Do well in exams. This clears your fundamentals and boosts CGPA.”

Interview (Round 2)

Tips before the interview (Aryaman):

• If you can show that you are interested in research and have seen what some of the labs work on, that will really set you apart. Mention what interests you, but be prepared to be asked some basic questions in that field.
• Note down all the things you want to mention in the interview - any internships you’ve done, research experience, or even a research paper that you read and that excited you. The interview is normally fast-paced, and you should not regret missing anything.
• It is obviously okay if you are not sure of your field of interest, but having a rough idea of what excites you / what you find cool is good enough.
• Be confident while answering, but not overconfident (you do not want to come off as arrogant). If you don’t know an answer, don’t give up - keep trying; professors tend to give hints and guide you to the right solution.

Online on MS Teams. There were 2 professors (P1, P2) in the meet, but their cameras were off - meanwhile Aryaman had worn a formal white shirt and a tie borrowed from a senior xD.

P1: “Tell us about yourself.”

Aside from blog: “Your introduction will literally decide everything. Stay composed and introduce yourself. Mention your motivation for joining IIIT, and your current research interest if you have thought about it. Also, mention courses in your current college that you liked studying. There is a good chance they will ask questions from these courses.”

Aryaman: “I am interested in the field of quantum computing. I enjoyed studying Data Structures, operating systems, linear algebra…”

P1 (interrupts): “Oh linear algebra! Tell us. Is the real number line a vector space?”

Super easy, I know. But I still managed to botch it up. Anxiety, I guess…

Aryaman: “Uhh, not sure.” (Nervous mumbling) “I know it is a field.”

P1: “Is the 2D space a vector space?”

Aryaman: “Oh yes it is. It is called R^2^ space.” (Realization hits) “Oh yes, even the real number line is a vector space. It is called R^1

P1: “Okay. What is a group.”

(Aryaman gives the definition, but it is not perfect.)

P1: “You mentioned field. What is a field?”

(Curses internally for mentioning fields. Gave a bad answer.)

P1: “Okay, leave that. You mentioned data structures. Describe Prim’s Algorithm.”

(Had made too many mistakes in the beginning, so confidence had taken a hit. Gave an okay-ish answer.)

P2: “Okay leave all that. You took Database Systems in college, right? What is the difference between natural join and inner join?” Follow-up: “Explain what the various joins do. Cross join, left outer, right outer, etc.”

(Had studied this almost 6 months before, so definitions were a bit rusty. Gave a decent answer, but not up to the mark.)

At this point, you might be wondering how Aryaman got in - but trust him, the rest of the interview is probably what changed it.

P1: “You mentioned quantum right? We have called a professor who works in that field here.”

(Some other voice starts speaking through P1’s laptop - P3.)

P3: “What is the significance of the wave function?”

Aryaman: “It describes the quantum state a system is in.”

P3: “Yes, but what else?”

Aryaman: “If you square it then you will get the probability density of the particle.”

P3: “But this function can take complex values.”

Aryaman: “Oh right! First, you find its magnitude and then square it.”

P3: “Okay. If ${\psi }_1$ is a solution to the Schrödinger wave equation and ${\psi }_2$ is another solution, then will ${\psi }_1+{\psi }_2$ be another solution?”

Aryaman: “Yes, it will be.”

P3: “Okay what if it was like ${\psi }_1+({\psi }_2{)}^2$?”

Aryaman: “No, it needs to be a linear combination of the valid solutions.”

P3: “Okay. Good. What else have you studied?”

Aryaman: “I have studied the basics. How qubits work, Dirac notation, quantum teleportation, basic key distribution, and algorithms like Deutsch-Jozsa and Grover’s search.”

P1: “Okay, thank you. Is there anything else you want to tell us?”

Aryaman: “I heard about IIIT’s quantum lab CQST, so I was hoping to join it if I got in.”

P1: “Okay, we will see.”

(what xD)

The interview ended there. Results came out within 5 days.

My Personal Lateral Entry Experience + Tips (IIITH) (Harith Y, 2025)

Background

July 22, 2024 was the date Harith found out about the Lateral Entry Scheme. Anjan Mahapatra bhaiya introduced it to him - “I’ll forever be thankful to him for that.” He was having lunch with his parents when he casually mentioned, “I’ll be joining IIIT Hyderabad next year.” They were confused and asked what he was talking about.

Harith was a CSE Dual Degree student at IIITDM Kancheepuram, simultaneously serving as Music Club coordinator, eSports Club coordinator, Tech Affairs Web Dev team member, Video Editing team member, Developers Club (formerly GDSC) coordinator, and Alumni Affairs content writer. Despite this packed schedule, consistent class attendance and well-organised digital notes (maintained purely out of love for subjects) carried him through - the GKP book he bought for preparation is still brand new.

Written Test Experience

• 3 weeks before the test, Harith had to manage his college’s Annual Fest, Vasisht $\times$ Samgatha ‘25 - being in 4 music bands and conducting a Prompt Engineering event. End-sems were also just 4 days after the LEEE test. Quite hectic days - and yet: no last-day revisions, no stress, no panic. Consistent note-making throughout the semester had already done the job.
• Exam hall anecdote: When Harith took his seat, there was a guy beside him, and a guy and a girl beside him (probably strangers to each other). Instead of focusing on the exam, they were talking. He gave them a few side-eyes; they stopped before he thought of reporting them. Still, it was a bit distracting.
• Format: Section 1 (Aptitude / Logical Reasoning, 30 q, 1 hour) + Section 2 (CSE 40 q + Math 15 q, 2 hours). Section 1 must be attempted first and cannot be accessed after the first hour - but you can switch to Section 2 early (cannot return).
• Math subsection: “very straightforward; almost direct, requiring minimal calculation by hand.” Theory-based; approach over rote.
• CSE questions: largely theory-based; syllabus spread fairly equally.
• Out-of-syllabus questions (OS, CN, DBMS): $\approx$3 DBMS + 2 CN + 3 OS = max 9 questions. Leaving all unattempted still allows clearing if core subjects are strong. If time permits, do cover them - useful at IIITH after joining too.
• Quant aptitude: “very easy, just logic” - no specialised preparation needed.
• English (reading comprehension) strategy (Harith): English questions consist of long paragraphs - it’s basically a “Search in the paragraph and answer the questions” type, so it won’t consume much time. Do not assume a keyword in the sub-question and search only for that keyword in a specific sentence
• the correct answer may depend on content from other parts of the paragraph associated with the same keyword. Read the entire paragraph; then locate the keyword in context; read all options before marking.
• “There’s no fixed syllabus for LEEE. It’s always the core topics you’ve covered in your previous institute until the end of the 2nd year.”
• Felt exam was easy; selected for interview on 12th May (2025 cycle).

Interview Preparation (9 focused days: 13th-21st May)

Subjects revised: C Programming, OOP in C++, Theory of Computing, DSA, Design and Analysis of Algorithms, Discrete Structures, Calculus, Differential Equations, Linear Algebra, Probability and Statistics, Digital System Design, Computer Organisation and Architecture, Database Systems.

Strategy: periodic rolling revision of all covered material + daily new topic. Analysed most commonly asked questions from past transcripts. Prioritised approach and problem-solving methodology over memorising correct answers. Relied heavily on MOOCs completed previously for a solid knowledge base.

Interview Experience

• Chose subjects: Probability, Data Structures, and Discrete Mathematics.
• Professors’ voices were faint - had to ask for repeats; normal, do not panic.
• Probability: 3 questions; could not answer the first; partially explained the second; struggled with the third. Thought he had failed.
• Discrete Mathematics: misinterpreted a relations question, explained his approach; professors politely corrected and continued. Admitted uncertainty on one sub-topic; professors immediately switched to a different topic.
• Linear Algebra + Real Analysis at the end; gave “convincing answers” despite imperfect recall; professors were giving hints he only recognised post-interview.
• Overall feel: “more like a discussion than an interview.”
• Selected on 4th June 2025.

Subject-wise Resources (Harith’s Interview Prep List)

• C Programming: GATE Smashers, Apna College
• OOP in C++: Apna College, Reference book
• DSA: GATE Smashers, Apna College, Apna Kaksha, Abdul Bari
• DAA: Abdul Bari, Lecture Series
• Discrete Structures: Lecture Series, GATE Wallah; Rosen book (highly recommended - problems used in competitive exams); Stanford free course on YouTube goes deep into every concept
• Calculus: JEE-level knowledge is enough; GATE Wallah
• Differential Equations: GATE Wallah
• Linear Algebra: Professor Dave + GATE Wallah + Rosen book
• Probability & Statistics: GATE Smashers, GATE Wallah
• Digital System Design: Neso Academy
• COA: GATE Smashers, CMPS 255
• Theory of Computing: GATE Smashers, Lecture Series
• Note: “The above references are only useful for revision. Learn from your own institute first. If not available, use NPTEL.”
• For the written test: GATE PYQs + understand how every algorithm works (be able to trace any algorithm). Understand theory, not definitions.

FAQs - Seats, Fees, Competition

• Seats: 15 max per branch (CSD and ECD) officially; 2025 CSD intake was 16 (panel’s discretion). Final decision based on interview, not written test.
• Competition: $\approx$5,000 write (CSD+ECD combined); $\approx$160 selected for interview; $\approx$30 get in total.
• Fees: $\approx$17-18 L total. B.Tech portion: $\approx$5L/year for 3 years = 15L. MS by Research: $\approx$1L/year. All figures including hostel and mess.
• CGPA: 8+ recommended, not mandatory. Half of the 2025 intake were from local/private colleges (no college-name bias). See CGPA Requirements for a full breakdown.
• Transcript deadline: If 4th semester results are delayed, 1 month grace period; can submit till $\approx$end of August. If only 3 semesters done by interview date, submit those 3 and provide 4th later - seat is not cancelled.
• 4th semester exam clashes: If LEEE exam clashes with your 4th sem exams, prioritise LEEE. Written exam is in the 2nd-3rd week of April.

FAQs - Exam Preparation Strategy

• When to start: Harith didn’t explicitly prepare - coursework covered it. Start treating every semester exam as a LEEE preparation round. This simultaneously boosts CGPA and LEEE readiness.
• GATE level?: Yes, GATE level is sufficient. The 2025 paper was “pretty easy” and entirely theoretical.
• Syllabus breadth: Cover all courses up to and including 4th semester at IIITH (not your own institute). Check IIITH CSE Curriculum page, exclude non-core courses (Sports, Arts, Value Education, Embedded Workshop, Electives).
• If DM/Digital Logic not taught at your college: Use Harith’s drive notes; Neso Academy for Digital Logic; Rosen book for Discrete Math.
• Limited time (DM + DAA + Digital Logic in 1 month): Difficult to master all three; do them in parallel and cover as much as possible. Prefer 1 month each.
• College consumes 9am-6pm: Use holidays/festivals to grasp concepts from external resources. Use winter/summer break to learn in advance or revisit unclear topics. “DO NOT POSTPONE ANY WORK GIVEN, THINKING THAT THERE’S TIME FOR THE DEADLINE EVEN THOUGH THERE MIGHT BE. SO WHENEVER YOU GET SOME WORK ASSIGNED, START DOING IT.”
• MOST IMP - treat sem exams as LEEE prep: While preparing for your semester examinations (internals, quizzes, mid-sems, end-sems), prepare for those exams as if you are giving LEEE. That way, your LEEE prep gets done simultaneously with your sem exams - boosting both your CGPA and your LEEE readiness.
• Plan B / Don’t ditch college: Your CGPA matters not only for LEEE, but also because - if you don’t crack LEEE - your Plan B is placement at your own college. Do as much as is required in every subject. Don’t ditch college entirely.
• DSA vs Web Dev: Focus on algorithms; web dev will not be of much use at IIITH (research-focused programme).
• Don’t use AI for SOP/application essay: Write it yourself; professors will likely not appreciate AI-generated content.
• Do projects matter for screening?: Unless a project is genuinely impressive (e.g., 100+ GitHub stars, or deployed and used by real users), putting it on your application doesn’t significantly change your shortlisting chances. Only mention projects you can discuss deeply in the interview.
• Do research papers help?: Yes - if the paper is Q3 or above (in Scimago journal ranking) or at a recognized conference. Below that threshold, it is unlikely to meaningfully improve your chances. A Q1/Q2 paper or a top-conference publication is a strong differentiator.
• Does competitive programming rank help?: LeetCode/Codeforces rank helps with the written exam and is useful in interviews, but a strong rank alone is not a significant screening differentiator - the interview assesses research potential, not just algorithmic speed.
• Instrumentation/AEIE students (ECD track): If you are from an instrumentation branch and have not studied Signals & Systems: in the written exam, skip signals questions and focus on topics you know. In the interview, proactively tell the panel: “I am from instrumentation; I have not studied signals. Can I get alternative questions?” The panel often accommodates this - a real aspirant (Tushar) got alternate questions after making this request.
• UGC fee rule: Colleges are not legally permitted to charge tuition fees for academic years during which a student is not enrolled or physically attending. If your college tries to charge full multi-year fees when you leave after 2 years for LEEE, you may have grounds to contest this under UGC guidelines. Verify with your college administration and the UGC norms applicable to your state/institution type before committing.

FAQs - Interview Conduct

• How to speak: English only. Treat it as a discussion. Think out loud. Ask for clarification. Be honest, open and confident in the interview. If you know the answer, explain step-by-step. If you don’t, ask for time and try; explain your approach even if wrong - professors will give hints or redirect. Be confident, but not overconfident (you do not want to come off as arrogant). Keep trying; professors tend to guide you toward the right solution.
• Self-introduction: Name, origin, research interests, labs at IIITH that caught your attention, professors you’d like to work under. No right answer - make sure your answer would satisfy you if you were the interviewer.
• Standing out: Research the IIITH website (labs, professors, publications) before the interview and express a specific, genuine interest.
• Prior research/internships needed?: Not mandatory; optional. Only mention projects/internships you’re confident answering questions about.
• Research interest in Cybersecurity: CSTAR Lab.
• Mock interviews: Practice with friends or family to improve confidence and communication.
• Interview duration range: Varies significantly - reported as short as 5-10 minutes for some candidates and 20-30 minutes for others. One candidate (AMITABH) had a 12-minute interview. Don’t read into the length.
• Interview day waiting time: Expect to wait for hours. One candidate waited from 9:00 AM to 4:15 PM before their slot. Bring something to keep yourself occupied and don’t let the wait affect your composure.
• CGPA buffer in interview: “If your CGPA is high (9, 9+), even if you mess up a little in the interview, I think they’ll let it go.” A strong CGPA gives slight margin for error; a low CGPA requires a proportionally stronger interview.
• LORs: The IIITH application form has no field to upload Letters of Recommendation - you cannot submit them even if you have them. LORs have no impact on selection. Prior connections with IIITH professors may marginally help.
• 12th board 75% cutoff: Community discussion did not confirm this as an enforced requirement. Do not treat it as definitive - check the official notification.
• Strong profile with low CGPA: Mention everything on the form - co-founded startups, internships at IIITH, exceptional competitive programming achievements. The form is your best opportunity to surface non-CGPA signals.
• CGPA honesty: Do not misrepresent your CGPA. “I suggest not to. I suggest just be honest.”
• Drops/gaps: A 4-year gap year is not disqualifying. Gaps between college years do not automatically make you ineligible.
• Prof. PK (PreCog): Helps with PhD referrals via LinkedIn for students interested in that path.

FAQs - Life at IIITH After Joining

• Quizzes: Both MCQ and written; offline; depends on the professor.
• Internals: Manageable if you attend classes and practice; some external reference helps. “Pay attention in class; they don’t ask anything other than what was mentioned in class.”
• Workload: Constant homework, assignments, mini-projects. Balanced if not procrastinated. Overwhelming in the first month; gets manageable with time.
• Classes: $\approx$8 classes $\times$ 1.5 hours / 6 days = 2 hours of class per day on average; free time expected to be used for self-study.
• Teaching style: “Profs here teach an adequate amount but at a relatively much faster pace and expect you to refer and practice on your own outside of class hours, whereas in my previous Institute, the profs used to explain everything in detail, make you solve problems, and do everything in class hours. So, there was very little need to go through the subject after the class hours.” - “Self-researching teaching methodology” vs the “spoon-feeding” style at IIITDMK; both have pros and cons.
• Adjusting: First few weeks are lonely; Imposter Syndrome is common; gets better as time progresses. “This feels like a game where you make a lot of progress, but then you revert to the checkpoint made at the beginning of the game. This jump gives you another opportunity in life to correct your 2nd year and do it properly this time, but also takes away 2 years of progress made in the previous Institute (Academically, 1 year).”
• Is it worth it?: “This jump is definitely worth it. Sacrifices have to be made when you’re joining the best research institute in India in terms of CSE.”

TLDR

• Research work and projects are not mandatory. Only mention them if confident.
• 8+ CGPA is recommended, not mandatory.
• Have fundamentals clear. Questions in both the test and interview are easy but theoretical. They want to know how you think, not if you know the answer.
• Follow your coursework: concentrate on classes, make well-organised notes, practise regularly. Do not learn by heart without understanding.
• Enjoy college life, but don’t neglect your studies. Be responsible.

Final Thoughts (Verbatim)

“This wasn’t merely a matter of changing institutes or seeking prestige. For me, it was about taking the road not travelled, one that was more in line with my love of learning, of research, of breaking the mould. I could have remained where I was, completed my degree, and let life tick along on default settings, but just that I wanted to do better than that. I wanted to discover. I had wished to earn my right at an institution like IIIT Hyderabad, where research is not a buzzword but a career path.

The interview, though not perfect (at all), was an eye-opener. It reminded me that honesty, approach, and clarity of thought are more important than getting everything right. I didn’t have all the answers, but I was honest about what I knew and tried to reason through everything. That’s what research is all about anyway: being curious, thinking deeply, and never pretending to know it all.

To anyone who’s considering LEEE or planning not to follow the herd, take the leap. It’s scary, yes. But it’s also worth it if you’re passionate. This journey has already changed the way I think, and I can’t wait to see how much more I grow in the years to come.

This was never just about clearing an exam. It was about choosing myself, every step of the way.”

Be who you needed

- Harith Y, 2025

Harith Y’s Personal Interview Questions (Memory-Based)

These are the actual questions asked in Harith Y’s LEEE interview, reconstructed from memory. The interview covered four topic areas:

A. Introduction & Academic Background

1. Could you please introduce yourself briefly?
2. Why do you want to leave your current institute and join IIIT Hyderabad?
3. What specifically attracts you to machine learning and research over an industry-oriented curriculum?
4. What is your current CGPA?
5. Which subjects or topics have you prepared for this interview?

B. Probability & Random Variables

1. What is a random variable?
2. What do you mean by a discrete random variable?
3. Can you name some commonly known discrete random variables?
4. Have you studied the geometric random variable?
5. Can you explain what the geometric random variable represents?
6. Do you know any probability distribution that has the memoryless property?
7. Suppose you repeatedly toss a fair coin - how many tosses are required, in expectation, to observe both a head and a tail?
8. What is the expected value of a geometric random variable?

C. Relations (Discrete Mathematics)

1. Have you studied relations in discrete mathematics?
2. Can you define reflexive and symmetric relations?
3. If two relations are both reflexive and symmetric:
4. What can you say about their intersection?
5. What can you say about their union?
6. If the relations are equivalence relations, what happens to their union and intersection?
7. Is the intersection of two equivalence relations always an equivalence relation?
8. Explain your reasoning.
9. Why does that property hold (or fail to hold)?

D. Algebra & Finite Fields

1. Have you studied finite fields?
2. Can you give an example of a finite field?
3. Why is the set of integers not a finite field?
4. What properties must a set satisfy to qualify as a finite field?
5. Is closure under addition and multiplication required for a finite field? Why?

Harith Y’s Interview Prep Plan (LEEE 2025)

Context note from Harith: “This is something I found in my LEEE ke time ka prep I planned for, once I got to know that I got selected for the interview. Whatever I have censored, don’t worry, either I have already mentioned before or it’s something personal to me. I am not censoring them because I want to gatekeep, but I did it so that you find your own interest and be honest and novel to yourself. And some questions have to be reflected by you and yourself, so answer them how you feel they should be, there’s no correct answer to such type of questions. You’re free to modify this to however you find useful. Plan stuff how you normally would do.”

Key point: The same schedule may not work for you. Take inspiration by reflecting on your own situation (Harith was fully free with no exams and reasonably strong in all concepts). Adapt the plan to your schedule, your tracking style, and your weaker areas.

Core Mindset for the Interview

• Know your research area: Be ready to explain which IIITH lab or research direction interests you and exactly why.
• Read at least one paper: Being able to discuss even one paper in your area shows genuine intellectual interest.
• Confidence and Honesty: If you do not know something, say so clearly - then explain what you do know around it.
• Strong problem solving + concept clarity: The panel tests depth of understanding, not breadth of memorisation.

”Tell Us About Yourself” Preparation Framework

• Your name, current college, and branch
• What specifically excites you about research (not just “I like CS”)
• 1-2 topics or projects you have explored beyond the curriculum
• Why IIITH specifically - name a lab, a professor, or a concrete research area
• An honest assessment of where you stand and what you want to learn here

12-Topic Study Checklist with Focus Areas

#	Subject	Focus Area / Key Subtopics
1	DSA	Space & Time Complexity
2	Discrete Mathematics	Graph Theory, Algorithms
3	DAA (Design & Analysis of Algorithms)	BFS, Dijkstra, Dynamic Programming, Sorting, Searching
4	Digital Logic	MUX, Flip-flops, Latches, Ripple & Look-ahead Carry Adders
5	COA (Computer Organisation & Architecture)	Assembly syntax & logic
6	C Programming	Recursion, Pointers
7	Probability & Statistics	Bayes’ Theorem, PDFs, Normalization
8	Linear Algebra	Basis, Orthogonal Matrices, Rank
9	Calculus & Differential Equations	Integration, differentiation, ODEs
10	OOP (Object-Oriented Programming)	new keyword, why the paradigm exists
11	Theory of Computing	DFA, $\epsilon$-NFA to NFA conversion
12	Database Systems	Normalization

12-Day Study Schedule (13 May - 24 May 2025)

The schedule ran from 13/05/2025 (post-shortlist) to 24/05/2025 (interview day). The structural rule on the plan: “previous day’s revision + current day’s topic” - each day began with a short review of the previous day before moving to the new subject.

Key checkpoints from the plan:

• 17/05 (mid-point revision pass): Excess-3 code, Real Analysis, P vs. NP, Pumping Lemma - a partial revision of early topics.
• 20/05 (second revision): Full revision pass - all earlier topics marked complete.
• 23/05 (final revision day): Full revision of all 13 subjects including AI/ML; mock interview questions reviewed; prior interview experiences reviewed; LinkedIn and GitHub profiles finalised; IIITH labs and faculty research pages reviewed.
• 24/05 (interview day): Final review of IIITH admissions application, lab and faculty context; final confidence check.

These are the motivational talking points Harith planned to articulate in his interview introduction:

• Intuition and practical understanding: Learning should produce genuine intuition, not just memorised formulas.
• Real learning over rote: A desire to understand why things work, not just what they are.
• Long-term goal: Contributing meaningfully to research in a related field; potential interest in PhD-level work to contribute to society.
• Acknowledged challenge: Time management - being honest about a real weakness shows self-awareness.

Non-Technical / Behavioural Questions to Prepare

Prepare answers to these from your own honest reflection. There is no single correct answer - copying someone else’s answers will come across as inauthentic to an experienced panel.

• Why IIITH? - Be specific: which lab, which research area, which faculty member and why.
• Tell me about yourself - Use the framework above.
• What is research? - Have a personal, genuine definition; not a textbook line.
• Research interests at IIITH - Which lab? Which problem? Why now?
• Team vs. working alone - How do you work best? Give a concrete real example.
• A real-life problem you want to solve - Identify something you genuinely care about.
• Strengths - Specific and honest; avoid generic answers like “I am a hard worker.”
• Weakness - Honest, with a genuine plan for addressing it.
• Future goals - PhD? Industry research? Startup? Have a real answer and a “why.”
• A challenge you faced - Describe what happened, what you did, and what you learned.
• Handling criticism - Show you can take feedback and update your thinking.
• Staying updated with tech trends - Name specific papers, blogs, or channels you actually follow.

After preparing each subject, verify:

1. Can you explain the concept in your own words to someone who has never studied it?
2. Do you know why the result or algorithm works - not just what it does?
3. Can you handle at least one non-trivial example or numerical without looking anything up?
4. Have you noted 1-2 things you are still genuinely unsure about in this topic (so you can say “I am less confident here” without being caught off guard)?

Arun Sivan (LCD ‘25) - Exam Reflection

“Honestly, the exam was quite good this time. While most questions were from the expected topics, time management was a bit tricky because some questions were logic-heavy and needed careful thinking. So I’d recommend not just studying the syllabus topics but also practicing application-based questions and previous years’ patterns if you can find them. Feel free to DM me on Facebook if you want more specific details. All the best for CSD ‘26!”

Arijeet Paul (LCD ‘25) - Detailed Preparation Tips

Mathematics:

• Start Discrete Mathematics early - it has almost no prerequisites. Use the NPTEL playlist by Ashish Chaudhary (IIIT Bangalore) for lectures and Kenneth Rosen for practice.
• Math section had 15 separate questions. This year it was very basic - JEE-type limits, maxima/minima, and engineering-math questions like eigenvalues, eigenvectors, Laplace transforms, and Z-transforms. “I did 14 out of 15 correctly.”
• For the interview, clarity on null spaces, vector spaces over a field, and theoretical linear algebra concepts matters - not just computation.
• Probability: MIT OCW course by John Tsitsiklis; Rosen for discrete probability.
• Laplace/Z-transforms were asked even though not explicitly in all prep guides - cover them.
• Unusual topics that appeared: stable marriage problem, four-colour theorem, two-colour problems.

DSA and Algorithms:

• MIT OCW 6.006 (primary), 6.046 (only MST part).
• CLRS (the standard reference). Algorithm by Dasgupta & Vazirani (highly recommended for conceptual clarity).
• Graph Algorithms: IISC Bangalore NPTEL course.
• Coursera UCLA DSA course.
• NPTEL course by Nitin Saxena (IIT Kanpur) - highly recommended for deep algorithmic understanding; assignments are challenging but excellent. “If you have time, they are quite hard but very good.”
• Practice variety across all subjects - do not exhaust all PYQs of one subject.
• Paper was not as hard as GATE PYQs but GKP books are a good base.

Exam Pattern (2025 experience):

• Approximate distribution: 30 CS questions + 15 Math questions (separate math section) + 30 aptitude questions.
• DBMS, OS, COA, Networks also appeared - “30% of the CS paper was out of syllabus.” Prepare these lightly as a safety net.
• Typical weightage order (varies year to year): DSA $>$ Discrete Math $>$ C Programming $>$ Digital Logic. However, in recent years the pattern has shifted unexpectedly - “this year’s paper was one big surprise.” Do not over-optimise prep around any single subject.
• “Math was easy. DSA had bottleneck spanning tree, graph questions overall were good. Aptitude is more of a speed test - practice it.”
• All questions are of equal marks.
• Prepare college courses thoroughly: “they somehow want to see how much you study in college as in interview also they asked me college-course-related questions.”
• Prepare alongside your college coursework. Starting a dedicated LEEE prep sprint 3-4 months before the exam will clash with your end-semester vivas and lab exams. Integrate GATE-level practice into your regular semester study from the beginning.
• GKP books + CLRS + Dasgupta were Arijeet’s primary resources.

Interview Advice:

• Check the Jeff Su LinkedIn videos on interview preparation.
• For DL revision: KnowledgeGate by Sanchit Sir - 11-hour one-shot on Digital Logic on YouTube; very good for revision.
• Focus on subjects taught in your 2nd year.
• “I was able to answer all but 1 question, and 1 question the interviewer was not happy with - ‘mehhh.’ They asked subject-related questions that were a bit tricky.”
• CGPA is not the only deciding factor; projects, GSoC, research are also strong points - highlight your strongest asset. “They directly asked me my GPA in the interview.”
• Join the Facebook LEEE group
• Reddit post with memory-based interview questions (shared post-results by a non-selected community member)

This document contains FAQ answers and over 24 documented interview experiences from the LEEE Facebook group community.

General FAQ Section

Q: What is the selection process?

A: Two-stage process:

1. Written Examination (GATE-level questions on CSE/ECE + aptitude)
2. Personal Interview at IIITH campus

Final selection is based primarily on the interview performance.

Q: What is the program structure?

A: You join as a 2nd-year B.Tech student in the B.Tech + MS by Research dual degree program. Total duration at IIITH: 4 years. You graduate with both B.Tech and MS degrees.

Q: What research domains are available at IIITH?

A: Robotics, ML, Computer Vision, Human Sciences, Natural Sciences, Computer Systems, Algorithms, Computational Social Sciences, Software Engineering, and many more.

Q: What do interviews typically focus on?

A: Mostly technical questions. They ask you to choose 1-2 subjects, and a professor asks questions in those areas. They may also ask about projects mentioned in your CV. They may ask about your research passion and why IIITH.

—

Interview Experiences 1-31 (2018-2023)

The following interview experiences were documented by the LEEE community on Facebook and compiled into this document. They span 2018-2023 and cover both LCD (CSD) and some ECD experiences.

Interview Experience #1 (circa 2018-2019)

The interview started with which college I was from and then moved to subjects. I chose Data Structures.

• What is a BST? How do you insert into a BST?
• What is the worst-case time for BST insertion?
• How can you avoid the worst case? (AVL trees / balanced BSTs)
• Explain BFS and DFS. When would you use each?
• Write pseudocode for BFS.
• What is the time complexity of BFS? Why?

Research motivation: Why do you want to come to IIITH? (They wanted to hear passion for research, not just “better college.”)

Interview Experience #2

Panel of 3 professors. Chose Discrete Mathematics.

• What is a relation? Give examples.
• What makes a relation an equivalence relation?
• Prove that the relation “has the same remainder when divided by 5” is an equivalence relation.
• What is a partial order? Give an example.
• What is the difference between a DAG and a tree?
• How many simple paths are there in a complete graph $K_n$?

Key observation: They started with easy questions and progressively increased difficulty. When I couldn’t answer a hard question, they moved on without penalizing me.

Interview Experience #3

Chose C Programming (chose this because I felt most confident).

• What is a pointer?
• What is the difference between int *p and int **p?
• What is malloc? How does it differ from calloc?
• Write a function to reverse a linked list in C.
• What is the output of this program? (showed a program with pointer arithmetic)
• What is a void pointer? When is it useful?

Research: Discussed interest in systems programming.

Interview Experience #4

Chose Algorithms and Probability.

• Explain merge sort. Write the code.
• What is the recurrence relation for merge sort? Solve it.
• Is merge sort stable? Is quick sort stable?
• Probability: Two dice are rolled. What is the probability that the sum is 7?
• Conditional probability: Given that at least one die shows 3, what is the probability that the sum is 7?
• Bayes’ theorem: A medical test has 99% sensitivity and 95% specificity. The disease prevalence is 0.1%. What is the probability that a positive test result means you have the disease?

Key moment: The Bayes’ theorem problem caught me off guard (counter-intuitive result). I worked through it slowly and they appreciated the careful reasoning even though I initially got the wrong answer.

Interview Experience #5

Panel of 4 professors. Duration: 30 minutes.

• From which college? What subjects studied?
• Chose Digital Logic as main subject.
• Difference between latch and flip-flop. (Very common!)
• What is a D flip-flop? Draw the circuit.
• Design a 3-bit synchronous counter using D flip-flops.
• What is the characteristic equation of a JK flip-flop?
• Simplify: $F=\sum (0,2,5,7)$ using K-map.
• Implement a full adder using half adders.

Interview Experience #6 (Early Online - COVID period)

Interview conducted online (Zoom/Teams) due to COVID-19 restrictions.

• Chose Data Structures.
• Explain a hash map. How does it handle collisions?
• What is the difference between open addressing and chaining?
• What is the load factor of a hash table? How does it affect performance?
• Online medium: Screen-shared my code and explained. The professors were patient with technical delays.
• Research motivation: discussed interest in ML research.
• Explain gradient descent in simple terms.
• What is overfitting? How do you address it?

Note on online interviews: During COVID years (2020-2021), all LEEE interviews were conducted online. The format was similar but candidates shared their screen or used a virtual whiteboard.

Interview Experience #7

Duration: 20 minutes. Chose Algorithms.

• Explain Dijkstra’s algorithm. What is its time complexity?
• Can Dijkstra handle negative edge weights? Why not?
• What algorithm would you use for negative weights? (Bellman-Ford)
• What is the difference between Prim’s and Kruskal’s for MST?
• Given a graph, run Prim’s algorithm from a given node (on paper).
• What is a topological sort? When can you apply it?

Interview Experience #8

Chose Discrete Mathematics and Linear Algebra.

Discrete Math:

• What is a group? What are the group axioms?
• Is $(\mathbb{Z},+)$ a group? Is $(\mathbb{Z},\times )$ a group?
• What is a subgroup?
• Is every subset of a group a subgroup?

Linear Algebra:

• What is the rank of a matrix?
• If a matrix $A$ has rank $r$, what is the dimension of its null space? (Rank-nullity theorem)
• What is an eigenvalue? If $A\mathbf{v}=\lambda \mathbf{v}$, what is special about $\mathbf{v}$?
• For a symmetric matrix, what can you say about eigenvalues and eigenvectors?

Interview Experience #9

Duration: 25 minutes. ML-focused interview (mentioned ML background).

• What is supervised learning? Give examples.
• What is the bias-variance tradeoff?
• How does a decision tree split nodes? What is information gain?
• What is regularization? Explain L1 vs. L2 regularization.
• What is gradient descent? What is stochastic gradient descent?
• Why is deep learning effective for computer vision?
• What is a convolutional layer? What does it compute?

Research: Expressed interest in computer vision research at CVIT. Professors asked about CVIT’s work.

Interview Experience #10

Graphs and Algorithms focus.

• Given a weighted directed graph, find the shortest path from source to all vertices.
• What is a strongly connected component?
• What algorithm finds SCCs? (Kosaraju’s / Tarjan’s)
• What is a bipartite graph? How do you check if a graph is bipartite?
• What is network flow? Explain the max-flow min-cut theorem conceptually.

Interview Experience #11

Chose Operating Systems (unusual choice - most choose DS/Algorithms).

• What is a process? What is a thread?
• What is the difference between process and thread in terms of memory?
• What are the Coffman conditions for deadlock?
• Explain the Banker’s algorithm.
• What is virtual memory? How does paging work?
• What is thrashing? How do you prevent it?
• Explain the page replacement algorithms: FIFO, LRU, Optimal.

Note: OS is a valid choice if you know it deeply. Don’t choose it just because it seems “easier.” The professors will probe deeply.

Interview Experience #12

Duration: 15 minutes (short!). Chose Probability.

• What is conditional probability?
• State Bayes’ theorem.
• Problem: A bag has 5 red and 3 blue balls. Two balls drawn without replacement. Probability both are red?
• Expected value problem.
• What is a Markov chain? (I said I wasn’t familiar; they moved on graciously.)

Takeaway: Even a short interview can lead to selection if the answers are strong and the research motivation is clear.

Interview Experience #13

Quantum computing interest.

• Chose Linear Algebra as subject.
• Eigenvalues and eigenvectors in depth.
• What is the spectral theorem?
• What is a unitary matrix?
• The professor noted that quantum computing relies heavily on linear algebra - asked if I was aware of this.
• Brief discussion on qubits and superposition (conceptual level).
• Research: Expressed interest in quantum information theory.

Note: You don’t need to know quantum computing for LEEE. But if you express interest, be prepared for follow-up questions.

Interview Experience #14

Duration: 20 minutes. Chose Data Structures and Digital Logic.

Digital Logic:

• Design a 4-bit ripple carry adder.
• What is propagation delay? How do carry lookahead adders solve it?
• Design a 3-to-8 decoder.
• Implement a 4-to-1 MUX using logic gates.

Data Structures:

• What is an AVL tree? How does it differ from a regular BST?
• Explain AVL rotations (LL, LR, RL, RR).
• What is the time complexity of AVL tree operations?

Interview Experience #15

Chose Algorithms and C Programming.

Algorithms:

• What is dynamic programming? How is it different from divide and conquer?
• Solve the 0/1 knapsack problem. Write the recurrence.
• What is the time and space complexity of your knapsack solution?
• Can you optimize the space? How?

C Programming:

• What is the output of this program? (tricky pointer program)
• What is a function pointer? Write an example.
• What is the difference between struct and union in C?

Interview Experience #16

Duration: 35 minutes (longer than average). Chose Discrete Mathematics.

• Define a graph. What is the difference between a simple graph and a multigraph?
• What is the handshaking lemma? Prove it.
• What is Euler’s theorem for graphs? (Necessary and sufficient conditions for Eulerian circuit)
• What is a Hamiltonian cycle? Is there an efficient algorithm to find one?
• This question led to a discussion of NP-completeness.
• What is P? What is NP? Is P=NP?
• What is a polynomial-time reduction?
• Why is the Hamiltonian cycle problem NP-complete?

Note: This interview went deep into computational complexity theory. The candidate was strong and the professors clearly enjoyed the conversation. Duration extended because both sides were engaged.

Interview Experience #17

Most recent pre-#18 experience. Chose Probability and Data Structures.

Probability:

• A fair coin is flipped 10 times. What is the probability of getting exactly 5 heads?
• What is the expected number of flips until the first head?
• Birthday problem: How many people are needed for a 50% chance that two share a birthday?
• Monty Hall problem: Explain it and solve it.

Data Structures:

• What is the difference between a min-heap and a max-heap?
• How do you build a heap in $O(n)$? Why is this faster than $n$ insertions?
• What is heapsort? What is its time complexity and is it stable?
• What is the difference between heapsort and mergesort? When would you prefer each?

Research: Interest in probabilistic algorithms and randomized data structures.

Interview Experience #18

4 interviewers. Interviewers first asked favourite subjects, then drilled on those.

• Best/average/worst-case time complexity.
• What is the best sorting algorithm and its time complexity?
• What do you get when you remove an edge from a tree? (A forest.)
• Define a tree; properties for a graph to be a tree. (Panel gave hints throughout.)
• Add an edge to a tree - what forms? (A cycle.) Prove it.
• How many cycles form when an edge is added? Prove.
• Point on which side of a line - solved on whiteboard with reasoning explained.
• Candidate mentioned Andrew Ng ML course on resume: What is a perceptron?
• Is logistic regression a classification or regression technique?

Interview Experience #19

Duration: $\approx$15 minutes (short). Panel was very friendly.

• Signals question (exact topic not recalled; answered, though wrong - panel prompted).
• Draw a 2$\times$1 MUX and implement AND and OR gates using it.
• Draw a 4$\times$1 MUX and implement its logic diagram.
• Three NOT gates in cascade, output fed back as input - what happens at output?
• Panel reviewed certificates; interview ended.

Key insight: Even a wrong answer with a confident attempt is acceptable. The panel prompts; stay calm.

Interview Experience #20 (Aakash, VIT Mumbai)

5 professors; candidate was the first interviewee of the day.

• “Why do you want to join IIITH?” - Answer emphasising desire for real engineering (not rote learning) visibly engaged the entire panel.
• “Did anyone else from your college qualify?” - Honest answer; don’t brag.
• SOP mentioned Profs. Amit Srivastava and Suryakanth (AI/NLP interest) - panelists noted all of them also work in AI; be prepared to justify specific choices.
• COA: Explain how a program moves from main memory to CPU (with diagram).
• Draw the graph for $n\log n$ - took many trials; panel helped, and everyone laughed at mistakes. Lightheartedness was a positive.
• Explain quicksort logic and why it is “quick.” Write it in code (no need for perfect code).
• Worst case of quicksort is $\Theta (n^2)$ - confirmed.
• “Your SOP says ML interest. Define machine learning in simple terms.” (Candidate blanked, thought for a minute in silence, then answered; was selected.)
• Final remark from professor: “I believe in human learning before machine learning” (wicked smile) - don’t read too much into sarcastic closing remarks.

Advice: Be confident and be yourself. Don’t put on a show. Bring humor and lightness.

Interview Experience #21

Duration: $\approx$25 minutes.

• What does research mean and how does it benefit society?
• Identify the curve for an equation (parabola, then changed to circle); find vertex.
• Puzzle: Perfectly cylindrical container filled with water. Without instruments, reduce volume to exactly half.
• Formal Languages (4th semester course): Design a DFA for a given table; convert an $ϵ$-NFA to DFA (shown how; stopped before fully solving on paper).
• Singly linked list with numbers 1-$n$; one missing. Find number of comparisons in worst case. Time and space complexity. What if sorted?
• What is a heap? Applications? Time complexities of operations?
• What data structures do you use in everyday life?
• Research interests; 10-year plan and how IIITH fits.

Advice: Clarify questions before answering. The professors will help with hints.

Interview Experience #22

• “Why IIITH?” - Answering “academics is good” is wrong. They want to hear love for research.
• “Which topic excited you most?” - This shapes the rest of the interview; choose something you can discuss in depth. Said “Divide and Conquer”; should have highlighted the visual recurrence tree intuition.
• Where is Divide & Conquer used? (merge sort, binary search, quicksort)
• Why use insertion sort over merge sort for cards dealt in sequence? (Merge sort is not in-place.)
• Write code for merge sort.
• Application of Fourier series - time domain to frequency domain.
• Given $n$ numbers, find duplicates; write code and exact time complexity with constants. Same problem with a linked list.
• Methods of mathematical proof: induction, deduction, contradiction…
• Prove $\sqrt{2}$ is irrational. (Candidate had no clue; panel showed it is more complex than repeated-decimal approach.)

Tips:

• Take relevant certificates, including Stanford/Coursera completions - professors from top institutes appreciate certificates from recognised platforms.
• “Which topic excited you most?” is the heart of the interview - it lets you steer. Always choose a topic you can sustain a deep conversation on.
• Never use a term you don’t know the meaning of.
• Keep trying a problem; never say “I can’t.” Let them stop you.

Interview Experience #23

• Research interest led to internship topic: “Algorithms to minimise Submodular Functions in Strongly Polynomial Time.”
• What is a modular function? What is a submodular function?
• Is probability modular or submodular?
• Domain and range of probability functions.
• Define a tree (graph theory definition).
• What is Functional Dependency? (DBMS)
• Favourite course: Data Structures.
• What is a stack and how does it differ from an array?
• Give examples where time complexity is good but space is bad, and vice versa. (Explained via two Sudoku implementations.)
• Linked list A$\to$B$\to$C$\to$D; you have the address of C only (no head pointer, singly linked). Remove C. (Solution: copy D’s data into C, then remove D.)
• Candidate mentioned ACM ICPC 2013 Amritapuri regional standing at the end.

Interview Experience #24

Topics: Cardinality (Discrete Mathematics), OS/Security.

• Chosen subject: Discrete Mathematics (picked by panel from candidate’s list).
• Cardinality of all real numbers vs. real numbers in $[0,1]$? (Candidate said “can’t compare two infinite values” - correct insight, wrong answer: both have the same uncountable cardinality by Cantor’s argument.)
• Cardinality of natural numbers vs. even numbers? (Again same cardinality - countably infinite; bijection $n\mapsto 2n$. Subsets of infinite sets can have equal cardinality.)
• OS/Security: Candidate mentioned a Windows XP exploit - asked: type of vulnerability, what is a vulnerability, how was the exploit executed, what made it threatening.
• Race conditions; semaphores; mutex.
• NLP basics (from SOP).

Key lesson: Do not assume cardinality of infinite sets follows subset rules.

Interview Experience #25

• Given a matrix traversal routine on paper; asked to write recursive code, explain logic, identify constraint on $n$ (eventually: $n$ must be a power of 2).
• Panel noted candidate’s college did security research; why come here? (Honest explanation that B.Tech students weren’t permitted access to that lab.)
• Panel pressed on CGPA dip - answered frankly (fluctuating interests).
• Favourite subject: Data Structures. Recursive factorial + complexity; recursive vs. iterative; which is better?
• Candidate explained research interest in real-world data structures on the web, UTF/Unicode challenges, unstructured data. Panel asked “what is unstructured data?” and ended the interview.

Interview Experience #27

Duration: $\approx$15 minutes.

• Mentioned Engineering Mathematics and complex numbers in courses.
• How do you say a function is analytic? (Cauchy-Riemann equations; what they mean.)
• Explain binary search; derive an algorithm for ternary search.
• Write the recurrence relation for ternary search; explain logarithmic running time.
• Why do you want to do research?

Interview Experience #28

4 teachers; threw multiple questions simultaneously; deliberate attempt to confuse candidate.

• Which subjects were taught last semester?
• What is the meaning of a probability distribution?
• Difference between a PDF and the standard distributions.
• What is a heap? How do you search in a heap?
• 8086 assembly: given a MOV statement - is it valid?
• Which field of research are you interested in?

Note: Some panels deliberately try to confuse. Maintain composure; answer one question at a time.

Interview Experience #29

Panel of 5. Started by reviewing badges and certificates on the file.

• Favourite subjects: Programming, OOP, Algorithms.
• Favourite topic: Compression algorithms $\to$ Huffman Coding. Explained on paper with example for $\approx$3 minutes.
• Why not Morse code instead of Huffman if Huffman is better? (Frequency dependency argument; fixed-length vs. variable-length encoding.)
• Time complexity of Huffman; panel disagreed repeatedly - candidate maintained his answer calmly. Lesson: stand your ground if you are correct and composed; don’t let the panel influence your answer when you’re right. Selected despite apparent disagreement.
• Code optimisation: for-loop faster than while; local variables faster than global.
• 2-wheel self-balancing robot project questions; went deep into manual robotics (candidate redirected to autonomous robotics focus, their actual domain).
• Conference volunteering mentioned as research motivation.

Interview Experience #30

• Confidence is the single most important factor. Convince the panel you are ready.
• Note: Intake is NOT fixed. The panel can select all interviewed candidates or none.
• After introduction, panel asks subjects covered across all semesters and your favourite among them. Choose a core subject and know it deeply.
• Interview revolves around 1-2 favourite subjects. Stay confident if the panel tries to confuse you.
• Use the paper with confidence; draw big, clear diagrams.
• Be honest about the 1-2 questions you don’t know the answer to.

Interview Experience #31

The interview narrows to 4 focus areas:

1. Favourite topics and interests (shapes the entire interview direction).
2. Favourite algorithm (be able to explain intuitively and analytically).
3. Last exam clarity (how well do you understand recent coursework?).
4. Motivation for research (must be genuine and specific).

Focus subjects: Data Structures, Discrete Mathematics, Computer Organisation, Digital Logic. Be well-versed in whatever you claim as favourites. Concept clarity is the theme, not memorisation.

Life at IIIT-H - Post-Joining FAQ

• Academic difficulty: If coming from a laid-back college, academics will feel gruelling - not difficult but expect a lot of work and programming.
• Catching up with 1st-year peers: IIITH introduces core CS in 1st year; you can audit (attend without credit) any course to catch up. Effort solves this.
• Research program structure: Choose a research advisor at end of 2nd year. Must complete 24 Thesis credits alongside courses. Advisors expect summer residency for research. Graduation depends on thesis completion - can exceed 4 years.
• Is it worth joining?: Only if you have genuine interest in learning or research (CS vision, robotics, AI, etc.). The research exposure at undergrad level is unparalleled in India. Joining purely for job prospects is not advised - workload is intense and graduation timeline is variable.
• LEEE vs. dropping for JEE: LEEE syllabus = your CS coursework; far less extra effort than JEE. LEEE carries significantly less risk because your current degree is your backup.

General Observations Across All Interview Experiences

Patterns from 24+ Interview Experiences

Universal patterns observed:

1. Every interview starts with: “Which college are you from?” and “What subjects are you comfortable with?”

2. The subject question determines everything. Your answer guides the next 20 minutes.

3. Progressive difficulty: Easy $\to$ Medium $\to$ Hard in each subject

4. Research motivation is universal: Every interview asks why you want to do research.

5. Professors are helpful: They give hints when you’re stuck. They’re not trying to fail you.

6. Honesty is rewarded: Saying “I don’t know” clearly, then attempting to reason, is respected.

7. What you mention, they’ll ask about: Every project, paper, or interest on your SOP is fair game.

Most common subjects chosen (in order):

1. Data Structures & Algorithms

2. Discrete Mathematics

3. Probability

4. Linear Algebra

5. Operating Systems

6. C Programming / Digital Logic (less common)

Most common “surprise” questions:

• Prove $\sqrt{2}$ is irrational (comes up in many DS/Discrete Math interviews)
• Cardinality of infinite sets (Cantor’s argument)
• Monty Hall problem
• What is NP-completeness?
• Graph theory puzzles (cylindrical container, line-point determination)

Additional Interview Experiences (2018-2021, detailed)

IE #1 (2018): Interview started after accidentally dropping a folder. Covered loop invariants on iterative factorial; linked list insertion under race conditions; swap operations on node contents. Key teaching moment: professors actively guide you toward the correct answer.

IE #2 (2018): 4 panelists. Tree distance problems using BFS; pairwise distance average calculation; P&C questions on functions; normal distribution and Bayes theorem with candidate-derived formula; group theory isomorphism; bijective functions requiring visual diagram representation.

IE #3 (2018): 4 tired panelists; student from IIITDM Kancheepuram. Questions: tree chromatic number (answer = 2); spanning tree definition; complete graph K4 subgraph enumeration; degree sequence and isomorphism; set theory subsets proof; linear algebra basics; conditional probability derivation; Bayes theorem application.

IE #4 (2018): GSOC project mentioned impressed the panel. Questions: C code swap variations (temp variable, arithmetic, pointer); matrix transpose with pointer arithmetic; two-coin Bayes theorem probability; birds-on-trees distance invariant puzzle; OS b-trees for disk data organisation; b-plus trees with leaf linkage. Interview duration: 26 minutes (confirmed via sign-in book). Student had attended Ramaiah IIT Study Circle.

IE #5 (2019): Array duplicate-finding with hash collision chaining; linear regression closed form (Newton descent, Hessian); matrix rank for rectangular $N\times M$ matrix ($=\min (M,N)$); column vectors. Student reported “way too bad” interview but was selected; CGPA was 9.65.

IE #6 (2019): First afternoon interview; blanked on “tell us about yourself”; lasted fewer than 20 minutes; chose Discrete Mathematics; interview cut short due to poor start. Highlights the importance of preparing a confident self-introduction.

IE #7 (2020, online): Deep Learning focus. Activation functions and non-linearity rationale; sigmoid range (0-1); CNN vs fully connected (parameter reduction rationale); black-box algorithms and CNN explainability; Cross Entropy Loss formula recall. Student disappointed by some gaps but was ultimately selected.

IE #8 (2020, online): Facebook friend storage via adjacency matrix/list; complete graph clique definition; Hamiltonian cycle traversal ($n!$ attempts given); stock exchange fast-computing, OS scheduling, and CN as real-world problems. Duration: 10-15 minutes.

IE #9 (2020, online): Competitive programming and ML hobbies. $N$-node complete graph: paths of length $k$; positive cycle for maximum reward path; reverse linked list vs. doubly linked list; dynamic programming and memoization.

IE #10 (2021, online): ACM ICPC regional qualification allowed candidate to bypass DS/Algo questions entirely. Java OOP questions (clone, new operator implicitly); virtual functions and pointer functions; functional programming vs. OOP; basis vectors in linear algebra and vector spaces; multiple inheritance in Java via interfaces. Duration: 17 minutes. Retrospective regret: “mentioned Java without full confidence.”

IE #11 (2021, online, 9 AM): 3 professors. Custom ROM project and bug fixes discussed; DFA and $ϵ$-NFA conversion exercises; graph distance problems; P&C and probability. Duration: fewer than 20 minutes.

IE #12 (circa 2020): Candidate chose Formal Languages and Automata Theory (uncommon choice). Questions covered: regular expressions, regular languages, Chomsky hierarchy classification; Computer Architecture - implement OR gate using a MUX; compare look-ahead carry adder vs. ripple-carry adder. Duration: 30 minutes. Tip from this candidate: “Confidence is the key. Think well before saying anything because that thing can be their next question and prepare a list of subjects in which you are comfortable.”

IE #13 (circa 2021): 2 hosts (cameras off). Adjacency matrix vs. adjacency list for storing social-network (Facebook-style) friend connections; space for adjacency list of a dense graph ($n$ nodes, each with $n/2$ degree connections). Key question: “What does multiplying the adjacency matrix $A$ with itself give?” - Answer: $A_{ij}^k$ counts the number of paths of length $k$ from vertex $i$ to vertex $j$. H2 gave hints throughout. Candidate had not prepared specifically for the exam and relied on strong college coursework. Confident interview overall.

IE #14 (2023): Chosen subjects: DLD and Discrete Mathematics. DLD: difference between flip-flop and latch (explained in depth); implement AND, NAND, and OR gates using a 4$\times$1 MUX; then redo using only 2$\times$1 MUX. Orthogonality of matrices (candidate admitted not knowing; panel did not stress). Discrete Math: basic number theory; 2-3 induction proofs (one straightforward, one solved with time, one only partially approached). Answered 5-6 of 9-10 questions fully. Duration: $\approx$20-25 minutes. Key lesson: always speak your thinking aloud even for incomplete answers. Be honest and polite above all.

IE #15 (2023): 2025 exam had two sections (general aptitude 30Q/60 marks/60 min; subject 50-55Q/100 marks/120 min with 25% negative marking). Interview: probability (mutually exclusive events vs. independent events) - candidate gave intuitive answer first (“if A occurs we are certain B has not, so they aren’t independent”) before working through the mathematical definition; discovered that the independence formula $P(A\cap B)=P(A)P(B)$ gives $0=0$ for a null event, creating an apparent contradiction. Quicksort $O(n\log n)$ derivation: couldn’t recall the analytical proof; panel hinted “it’s divide and conquer” - leading to the recurrence relation. Hash table with chaining for 1000 numbers with one duplicate pair: candidate couldn’t find an $O(n)$ solution. Duration: 17 minutes; selected.

IE #17 (undated): Research at IIITH works through labs; dual degree students choose a lab by their 4th semester. Research domains active at IIITH: robotics, ML, computer vision, human sciences, natural sciences, computer systems, algorithms, computational social sciences, software engineering. Interview format: mostly technical (choose 1-2 subjects); CV projects discussed; passion and research interest asked. Advice: “Focus on your exam first, worry about the interview later.” Join the IIITH LEEE Facebook page for previous interview experiences (https://www.iiit.ac.in/research-centres/ for lab list).

Exam History

• Format changed from subjective to MCQ in 2018.
• 2018 approximate cutoffs: 32/75 ECD, 37/75 CSD.
• 2019-2020: No answer keys released publicly; only interview invitations sent.
• 2023 cycle: Written exam May 6; interview May 27 (21-day prep window between results and interview).

Additional Books Recommended

See Resources from Harith’s Blog (Part VIII) for the consolidated book list covering C, DSA, Algorithms, Probability, Linear Algebra, Discrete Math, and Engineering Mathematics.

GPA Calculation (Cumulative)

GPA is computed as: [ \text{CGPA} = \frac{\sum_{i} (\text{grade}_i \times \text{credits}i)}{\sum{i} \text{credits}_i} ] It is not the simple average of semester GPAs. This distinction matters when reporting your CGPA in the application.

Key Interview Conduct Points

• Use the paper actively; professors explicitly encourage writing to see your thought process.
• On difficult questions: explain your approach aloud even if wrong.
• Topics mentioned on resume/SOP become interview material - prepare deeply for anything you mention.
• Confidence is valued; arrogance is penalised. If stuck, ask for hints freely.
• Research interest assessment: rough ideas about what excites you are sufficient; honesty about uncertainty is acceptable.
• “Why IIITH?” answer should emphasise research environment and academic quality, not just “better college”.
• Some panels ask why you are leaving your current college - have a thoughtful, positive answer ready.

Related

• ecdled-experiences