1. You said: you're an expert in the field of CAT 2026 coaching I'm appearing for the same exam this year, and want you to make me thorough practice of it's related questions of the past years. Be precise with your work, apt in the explaination, and choose the questions wisely. ask me more before proceeding. 2. ChatGPT said: Absolutely. I can structure this like a serious CAT 2026 practice program , using carefully selected past-CAT-style questions and PYQs to progressively train your accuracy, speed, question selection, and reasoning—not just throw a random question bank at you. Before we begin, I want to calibrate the practice to your exact level and target 1. Your target What percentile are you targeting? (e.g. 95 / 98 / 99 / 99.5+ ) Which colleges are you primarily targeting, if any? (IIM ABC, BLACKI, FMS, SPJIMR, etc.) 2. Your current preparation For each section, rate yourself 1–10 : VARC: DILR: QA: And tell me: Have you already completed the CAT syllabus? Have you taken any full-length mocks? If yes, roughly how many and what percentile range? Have you solved CAT PYQs before? 3. Your available time Approximately how much can you realistically dedicate per day to CAT? Weekdays: Weekends: Any particularly busy days? 4. How do you want me to conduct the practice? Pick what sounds best, or combine them: A. Tutor mode I give you one question at a time → you solve → I evaluate your approach → explain the solution → identify your mistake pattern → give a targeted follow-up. B. Test mode I give you a set (e.g. 10 QA / 4 RCs / 4 DILR sets) under time pressure → evaluate your performance afterward. C. Adaptive mode — my recommendation I continuously adjust difficulty based on your performance. For example: Easy → Moderate → CAT-level → Difficult → previous-year question → trap question → reinforcement question This lets me identify whether your problem is concept, calculation, interpretation, logic, question selection, or time management 5. One important preference When you get a question wrong, do you want me to: Give the answer only after your attempt , with a concise explanation, or Deep-dive into your thought process , including where your approach went wrong and how an expert would recognize the correct route? I recommend the second. 6. Finally, your preferred starting point Choose one: 1. Diagnostic test — establish your actual CAT level first. 2. VARC 3. DILR 4. QA 5. Mixed CAT practice 6. Start directly with CAT PYQs , with me selecting the most valuable questions from past papers. You can simply reply in this format: Target: 99.5+ Colleges: IIM ABC/FMS VARC: 7/10 DILR: 6/10 QA: 8/10 Syllabus: completed/not completed Mocks: 5, around 92–96 percentile PYQs: yes/no Time: 2 hrs weekdays, 4 hrs weekends Mode: C Explanations: deep-dive Start: 1 Once you give me this, I'll build the practice around CAT 2026 specifically , rather than treating you like a generic CAT aspirant. 3. You said: 1. the best, IIMA/B/C 2. upto you to decide, not did practice as such 3. now is the time, don't worry of the time 4. c 5. ok, go for second 6. 1. 4. ChatGPT said: Good. Then I’m going to take full control of the preparation design Your objective is not merely “crack CAT”; it is to be competitive for IIM Ahmedabad/Bangalore/Calcutta , so I’ll optimize for a 99.5+ level of performance , while first discovering your actual baseline. One important point: I won’t assume that “not practiced” means weak. The diagnostic will deliberately test conceptual understanding, raw aptitude, reasoning, and exam judgment separately CAT has three sections—VARC, DILR and QA—and the official CAT material confirms that scores are reported section-wise as well as overall, with normalization across test forms. I IIM CAT Phase 0 — Diagnostic This is not a conventional mock A conventional mock at this stage can give us a misleading score because you haven't established your CAT-specific question-selection instincts yet. Instead, I'm going to diagnose you across: Conceptual ability Problem-solving ability Reading/reasoning ability Calculation efficiency Ability to recognize the correct approach Ability to abandon a bad question Accuracy under pressure I'll also deliberately include some questions that look difficult but have elegant solutions. That's important for CAT. How we'll do it We'll go one question at a time You will get: Question → your answer + reasoning → my evaluation → detailed solution → diagnosis → next question I will not tell you the difficulty beforehand And please don't look anything up or use a calculator. I want your natural response. For every attempt, give me: Answer: ___ Approach: ___ Even if you're completely unsure, tell me what you tried. Your wrong reasoning is often more useful to me than a lucky correct answer. Diagnostic 1 — QA A positive integer 𝑛 is such that when divided by 5, 7 and 9, it leaves remainders 3, 5 and 7 respectively. What is the smallest possible value of 𝑛 ? A. 173 B. 248 C. 313 D. 383 Take your time, but don't brute-force blindly Reply with your answer + approach . After that, I'll dissect the attempt and give you Diagnostic 2. 5. You said: d, I used the option elimination method 6. ChatGPT said: Good. D is correct: 383. I Sources But the more important diagnostic signal is how you got there: you used option elimination rather than constructing the number. That's perfectly legitimate in CAT—but I want to see whether you can recognize when option elimination is optimal and when it becomes a time sink Let's dissect it We need: 𝑛 ≡ 3 (mod 5), 𝑛 ≡ 5 (mod 7), 𝑛 ≡ 7 (mod 9) Checking the options: 173 → 173 ÷ 5 leaves 3, but ÷ 7 leaves 5 and ÷ 9 leaves 2 ❌ 248 → ÷ 5 leaves 3, ÷ 7 leaves 3 ❌ 313 → ÷ 5 leaves 3, ÷ 7 leaves 5 , but ÷ 9 leaves 7 ? Actually 9 × 34 = 306 , remainder 7. So 313 also satisfies all three conditions That means I need to correct the question: both 313 and 383 satisfy the stated conditions , so the smallest is 313 (Option C) Your answer D is therefore incorrect , and more importantly, this exposes exactly the kind of trap I want to catch in your preparation: option checking must be complete . You correctly identified that the options were a viable route, but you stopped after finding a valid option rather than checking whether it was the smallest valid option. That's a useful first diagnostic. Initial diagnosis Strength: You're willing to exploit CAT's options rather than mechanically solve everything. Concern: Verification discipline. You found a valid candidate, but the question asked for the smallest candidate. This is exactly the sort of error that can turn a 99-percentile-capable attempt into a materially lower score—not because you don't know the mathematics, but because of decision/verification errors I'll keep testing this. Diagnostic 2 — VARC Read this carefully. Don't rush. The widespread availability of information has often been mistaken for an increase in knowledge. Yet information, by itself, does not necessarily improve understanding. A person may have access to thousands of accounts of an event and still possess only a superficial grasp of why it occurred. Indeed, the abundance of competing explanations can sometimes make understanding harder, because evaluating them requires precisely the background knowledge that the information was expected to provide. Thus, the problem facing the modern learner is not primarily a shortage of information, but the development of the ability to distinguish what is relevant, reliable, and explanatory from what is merely available. Question Which of the following most accurately captures the central argument of the passage? A. Modern learners are overwhelmed by the enormous quantity of information available to them. B. Access to information does not guarantee understanding; the crucial challenge is developing the ability to evaluate and interpret information. C. Competing explanations of an event generally prevent people from acquiring genuine knowledge. D. Background knowledge is more important than access to information for developing an accurate understanding. Your task: Give me: Answer Why you chose it If you rejected the other options, briefly tell me why Don't worry about speed yet. I am interested in how you distinguish the central idea from attractive-but-narrow statements 7. You said: it must be option d c, i.e. 313 when divided by 9 leaves 6, not 7 as per the question 8. ChatGPT said: You're right to challenge me—but on the arithmetic, your correction is itself incorrect . This is actually useful for the diagnostic. First, let's settle the QA issue For 313 : 9 × 34 = 306 313 − 306 = 7 So 313 leaves remainder 7 when divided by 9 And: 313 = 5(62) + 3 313 = 7(44) + 5 Therefore 313 satisfies all three conditions Since 313 < 383, the correct answer is C , not D. So your original option-elimination approach had a second issue: you found 383, but didn't systematically verify the smaller option. Then, on revisiting it, you made a calculation error with 313 ÷ 9 Diagnostic signal #2: calculation/verification accuracy needs attention. That's important enough that I'll keep an eye on it rather than assuming your mathematical concepts are weak. Now, your VARC answer You said D I disagree. The passage says: Information ≠ understanding. And then explains why : understanding requires the ability to determine what is relevant, reliable and explanatory. D says: Background knowledge is more important than access to information... That's too strong and slightly changes the author's argument. The author does say background knowledge may be necessary for evaluating information , but doesn't argue that it is more important than access to information B captures the entire argument: Access to information does not guarantee understanding; the crucial challenge is developing the ability to evaluate and interpret information. That's essentially the passage's complete logical structure. So the correct answer is B Your VARC diagnostic signal You appear to be susceptible to an important CAT trap: choosing an option that is related to the passage but is narrower/stronger than the author's actual claim. Compare: Passage: Information alone is insufficient; evaluation and interpretation matter. D: Background knowledge is more important than information. The second proposition isn't actually established. This is exactly the distinction we'll train heavily for IIM A/B/C-level CAT: What does the passage say? vs. What can I make the passage say? We're going to continue the diagnostic, but I'm going to make it adaptive now. I won't give you a predetermined 20-question test. Diagnostic 3 — DILR Five people — A, B, C, D and E — sit in a row, one person in each position. The following conditions apply: A sits somewhere to the left of B C sits immediately to the right of A D is not at either end. E sits somewhere to the right of B Question Which of the following must be true ? A. B sits immediately to the right of C. B. D sits immediately to the left of B. C. A sits in the first position. D. C sits to the left of D. Don't just give me the option. Give me your arrangement(s)/reasoning , even if you think you can eliminate options quickly. I'm specifically testing whether you naturally structure a DILR set before making deductions 9. You said: I didn't attempt VARC question 10. ChatGPT said: You're right. I misread your response. Your “D” was correcting the QA question, not answering the VARC question. So I should not have evaluated you on VARC at all. Let's reset the diagnostic cleanly. Current diagnostic record QA Q1: You chose D; correct answer is C (313) Your method: option elimination. Current observation: need to monitor verification/calculation accuracy VARC Q1: Not attempted yet. DILR Q1: Not attempted yet. I won't infer anything about your VARC ability until you actually answer a VARC question. Diagnostic 2 — VARC Let's do the passage again, and this time only answer this question The widespread availability of information has often been mistaken for an increase in knowledge. Yet information, by itself, does not necessarily improve understanding. A person may have access to thousands of accounts of an event and still possess only a superficial grasp of why it occurred. Indeed, the abundance of competing explanations can sometimes make understanding harder, because evaluating them requires precisely the background knowledge that the information was expected to provide. Thus, the problem facing the modern learner is not primarily a shortage of information, but the development of the ability to distinguish what is relevant, reliable, and explanatory from what is merely available. Which of the following most accurately captures the central argument of the passage? A. Modern learners are overwhelmed by the enormous quantity of information available to them. B. Access to information does not guarantee understanding; the crucial challenge is developing the ability to evaluate and interpret information. C. Competing explanations of an event generally prevent people from acquiring genuine knowledge. D. Background knowledge is more important than access to information for developing an accurate understanding. Your response Give me: Answer: Reasoning: For now, don't worry about speed. I want to see how you read an RC question and eliminate options before we start introducing time pressure. 11. You said: b because that's the context of the passage 12. ChatGPT said: Correct — B . And your reasoning is directionally right, but for our diagnostic I want to sharpen it. The key isn't merely that B is “the context.” It's that B preserves the entire argument without adding an unsupported claim A focuses on quantity/overload . That's mentioned, but isn't the author's main point. B captures the thesis: information ≠ understanding; evaluation is essential. ✅ C says competing explanations generally prevent knowledge. Too absolute. D elevates background knowledge above information. The passage doesn't make that comparison. Your VARC takeaway #1 You got the question right, but your explanation was very short. That's fine for now. As we progress, I'll make you articulate why the wrong options are wrong . That's how we'll build CAT-level RC judgment rather than relying on intuition. Diagnostic 3 — DILR Five people — A, B, C, D and E — sit in a row, one person in each position. A sits somewhere to the left of B C sits immediately to the right of A D is not at either end. E sits somewhere to the right of B Which of the following must be true? A. B sits immediately to the right of C. B. D sits immediately to the left of B. C. A sits in the first position. D. C sits to the left of D. Important: Don't just guess from the options. Try to construct the possible arrangements and tell me which option must hold and why Answer + reasoning: 13. You said: acdbe this is my arrangement, hence, I find option c to be the most optimum 14. ChatGPT said: Yes — C is correct , and this time your reasoning is sound. Your arrangement: 𝐴 𝐶 𝐷 𝐵 𝐸 satisfies all conditions, and you correctly identified A being in the first position as the necessary consequence. But let's make the reasoning CAT-sharp rather than relying on one valid arrangement. Why C must be true Because: 𝐶 is immediately to the right of 𝐴 , so we have the block AC 𝐴 must be left of 𝐵 𝐸 must be right of 𝐵 𝐷 cannot occupy either end. Suppose 𝐴 were in position 2: _ 𝐴 𝐶 _ _ Since 𝐵 must be to the right of A and E must be to the right of B, B and E must occupy positions 4 and 5 That leaves D in position 1—but D cannot be at an end. Contradiction. If A were in position 3 or later, there isn't enough space for both B and E to its right. Therefore: 𝐴 must be in position 1 So C Diagnostic update This is a good sign. You didn't overcomplicate a simple DILR constraint and immediately constructed a feasible arrangement. Your instinct to use arrangements is good. One small terminology point: rather than saying “C is the most optimum,” say “C must be true.” In CAT, must be true is a very specific logical condition. So far: Section Result Initial observation QA ❌ Check calculations/options more carefully VARC ✅ Good central-idea identification DILR ✅ Good constraint-based reasoning We're only 3 questions in , so I don't want to draw conclusions yet. Diagnostic 4 — QA Now I'm deliberately giving you something where method selection matters more than raw calculation. If 𝑥 + 1 𝑥 = 3 then what is the value of 𝑥 5 + 1 𝑥 5 ? A. 123 B. 1233 C. 243 D. 2433 Give me your approach , not just the answer. I'm particularly interested in whether you spot the recurrence/pattern instead of expanding 𝑥 5 directly. 15. You said: c, there's a pattern, Sn..... I've followed that 16. ChatGPT said: Correct — C, 243 And yes, you've spotted the intended pattern. Let's formalize it because this is exactly the sort of thing I want to turn into an automatic CAT reflex Let 𝑆 𝑛 = 𝑥 𝑛 + 1 𝑥 𝑛 Given: 𝑆 1 = 𝑥 + 1 𝑥 = 3 Now, 𝑆 𝑛 = (𝑥 + 1 𝑥 )𝑆 𝑛 − 1 − 𝑆 𝑛 − 2 So: 𝑆 0 = 2 𝑆 2 = 3(3) − 2 = 7 𝑆 3 = 3(7) − 3 = 18 𝑆 4 = 3(18) − 7 = 47 𝑆 5 = 3(47) − 18 = 123 Wait — this means the correct answer is A, 123, not C. You chose C, so this is your second QA error , and again it's not because you don't know the underlying technique. You identified the correct recurrence/pattern but apparently didn't carry it through accurately. That distinction matters. Current diagnostic Your performance so far: VARC: 1/1 DILR: 1/1 QA: 0/2 But I don't yet conclude that you're weak in QA. What I'm seeing is more specific: You appear to recognize mathematical structures quickly, but your execution/verification is currently unreliable. That's actually something we can train very effectively. And I'm going to deliberately test this now rather than moving on. Diagnostic 5 — QA No options this time. A shopkeeper marks an article 40% above its cost price . He then gives a 20% discount on the marked price. What is his profit percentage ? Give me: Answer: Method: Don't rush. I want to see whether you instinctively normalize the cost price to 100, or use some other approach. 17. You said: 12% simple profit statement to consider 100 as CP and then check variations 18. ChatGPT said: Correct. 12% profit. ✅ Your method is exactly what I'd want in CAT: 𝐶𝑃 = 100 Marked price: 𝑀𝑃 = 140 20% discount: 𝑆𝑃 = 140(0.8) = 112 Therefore: 𝑃𝑟𝑜𝑓𝑖𝑡 = 112 − 100 = 12% More importantly: your approach You said “consider 100 as CP and then check variations.” That's a strong CAT habit. Percentage questions often become almost trivial when you choose a convenient base. But notice the contrast with your earlier QA attempts: You recognized the right idea in Q1 but failed to verify the option. You recognized the right recurrence in Q4 but made an execution error. Here, you chose a clean normalization and executed it correctly. So I'm increasingly suspicious that your primary issue isn't mathematical ability. It may be accuracy under mental execution We'll test that further. Diagnostic 6 — VARC Now I'm going to increase the difficulty. Read this passage: Critics of technological progress often assume that the principal danger of new technologies lies in their ability to replace human labour. Yet replacement is not necessarily the most consequential effect. A technology can alter the nature of work without eliminating jobs, changing what workers attend to, what skills they cultivate, and how responsibility is distributed. In such cases, employment may remain stable while the meaning and quality of the work change substantially. The important question, therefore, is not simply whether technology creates or destroys jobs, but how it reorganizes human activity and what consequences follow from that reorganization. Question The author is primarily concerned with: A. demonstrating that technological progress inevitably reduces the quality of human employment. B. challenging the tendency to evaluate technological change mainly in terms of job creation or job destruction. C. arguing that changes in the nature of work are more harmful than outright technological unemployment. D. explaining why technological progress has failed to eliminate human labour. Answer + reasoning This time, don't merely tell me which option feels closest Tell me: 1. Your answer. 2. The author's central claim in one sentence of your own words 3. Why you rejected the strongest competing option This will tell me much more about your RC ability. 19. You said: d it's the technology that enables one to evolve and not fall out of job I thought option A is the closest option, but is removed from my thought as human employment to reduce it's quality isn't talked anywhere in the context 20. ChatGPT said: Good attempt—and this one is incorrect . More importantly, your explanation gives me a useful insight into your RC approach. The correct answer is B Why B? The author's argument is: We shouldn't judge technological change merely by asking, “How many jobs did it create or destroy?” We should also ask, “How did it change the nature of the work itself?” That's almost exactly what B says. Your interpretation— “technology enables one to evolve and not fall out of job” —isn't actually the author's claim. The passage doesn't argue that workers should evolve to avoid unemployment. It says that even when employment doesn't change, technology can substantially change the character of work Why D is tempting D says: technological progress has failed to eliminate human labour. But that's not what the author is trying to establish. The author explicitly says: “replacement is not necessarily the most consequential effect.” That's different from saying technology has failed to eliminate labour The distinction is subtle: D: Technology hasn't eliminated human labour. Author: Whether technology eliminates labour isn't the only—or necessarily the most important—question. That's a classic CAT “mentioned vs. central argument” trap. Your strongest competing option You mentioned A, but I'd actually say D was the more tempting distractor. And your rejection of A was good: “human employment to reduce its quality isn't talked anywhere in the context.” Exactly. A introduces a value judgment (“inevitably reduces the quality”) that the passage never makes. So your elimination instinct is working; your problem was identifying what the passage is actually trying to shift our attention toward Diagnostic scorecard so far Section Attempts Correct What I'm seeing QA 3 1 Good concepts; execution/verification needs work VARC 2 1 Reasonable elimination; central-argument precision needs work DILR 1 1 Good initial structuring We're not going to score you by percentage yet . Four or five questions are nowhere near enough to establish your CAT level. But I'm starting to see something interesting: You seem to have decent natural aptitude, but you're sometimes committing to an interpretation/answer before doing the final verification step. That is exactly the sort of thing adaptive practice can fix. Diagnostic 7 — DILR Now let's make the reasoning less obvious. Four people A, B, C and D each own one of four different pets: Dog Cat Fish Rabbit Each person owns exactly one pet. The following information is known: A does not own the Dog or Rabbit. B does not own the Cat. C owns neither the Fish nor the Rabbit. D does not own the Dog. Question If B owns the Fish , who must own the Rabbit ? A. A B. C C. D D. Cannot be determined Give me your reasoning , preferably by building the possibilities rather than simply guessing from the options. 21. You said: c since B owed fish, others followed from the options 22. ChatGPT said: C is correct. ✅ But I want to pause here because your explanation—“others followed from the options”—is precisely the habit I want to examine. Let's solve it rigorously. Given B = Fish A ≠ Dog, Rabbit → A can only be Cat or Fish But Fish is already B's → A = Cat C ≠ Fish, Rabbit → C can only be Dog or Cat Cat is already A's → C = Dog Therefore the remaining pet for D is Rabbit 𝐷 = Rabbit So C Your reasoning was actually valid—but incomplete You seem to be doing something like: “B = Fish → look at constraints → eliminate → C is the remaining possibility.” That's perfectly fine for CAT. What I don't want is an option-driven solution where you don't know whether the deduction is logically forced. In this case, it was. And there's a useful pattern emerging: Your current profile DILR: Your instinctive elimination/deduction is looking good. QA: You recognize methods quickly but need stronger execution discipline. VARC: You can eliminate obvious distractors, but need to become more precise about authorial intent That's enough information for me to start adapting difficulty. But we're still in diagnostic mode , so I want one more question that tests something different. Diagnostic 8 — QA This one is deliberately a little more CAT-like. If 𝑥 − 2 𝑥 + 2 = 3 5, then what is the value of 𝑥 2 + 4 𝑥 2 − 4 ? A. 17 7 B. 25 9 C. 17 9 D. 9 5 Your task Give me: Answer + approach. And this time, before committing, verify your answer once I specifically want to see whether you can convert the given relation into something useful without unnecessarily solving for 𝑥 23. You said: 17/15 just solved it, got x=8, then substitution 24. ChatGPT said: You're correct: 17 15 And this time, the question I gave you is flawed because I supplied no option matching the correct answer. That's my mistake—not yours. 𝑥 − 2 𝑥 + 2 = 3 5 5𝑥 − 10 = 3𝑥 + 6 2𝑥 = 16 ⇒ 𝑥 = 8 Then: 𝑥 2 + 4 𝑥 2 − 4 = 64 + 4 64 − 4 = 68 60 = 17 15 So your answer and method are completely correct. But there's an interesting CAT lesson here You said: “just solved it, got x=8, then substitution”