Sample For Representation Regarding Answer Key: Fill & Download for Free

NEET Result 2019 is expected to be released on June 5, 2019. The minimum NEET cut off marks to qualify the examination will be 50% (for General category) & 40% (for Reserved category). The results will be available at the official portal i.e. National Testing Agency in the form of score card. The scorecard for NEET 2019 contain candidates’ All India Rank (quota and category), overall & sectional marks and percentile. The counselling procedure will be started after the result declaration.Latest Application Forms 2019APPLY FOR UG/PG COURSES AT BENNETT UNIVERSITYTIMES OF INDIA GROUP | SCHOLARSHIP UPTO 100%*7 DEDICATED RESEARCH LABS | 10 CR+ RESEARCH GRANTSADMISSIONS 2019 OPEN AT IFIMAwarded AACSB Accredition24 YEARS OF EXCELLENCEADMISSIONS OPEN AT CMR UNIVERSITY, BANGALOREApply Now for UG & PG ProgramsREGISTER NOW FOR CMRUAT 2019 | CMR UNIVERSITY ADMISSION TESTADMISSIONS 2019 OPEN AT CHANDIGARH UNIVERSITYUNIVERSITY WITH BEST PLACEMENT RECORDSDISCOVER EXPERIENTIAL LEARNING APPROACHThis year, NTA has scheduled the examination for May 5 in offline mode. Candidates qualifying the examination will have an opportunity to grab a seat into approximately 65,000 MBBS and 25,000 BDS seats in medical and dental colleges across India. There will be no provision of re-evaluation/re-checking of the answer sheets. After the declaration of cutoff, an All India Merit List will be prepared by NTA, containing names of qualified candidates with their corresponding NEET Rank.NEET CutoffHow to Calculate NEET Score & RankNEET 2019 CounsellingNEET Answer Keys 2019#Latest UpdatesCandidates from UR category aged above 25 years and those with Biology as an additional subject may also apply for the exam.NEET Merit List will now be used to allot seats in veterinary colleges for undergraduate courses like BVSc. and Animal Husbandry under 15% All India Quota.NEET Result will be applicable for admissions to all medical colleges of India except for AIIMS and JIPMER, which conduct their own medical entrance exams.NEET 2019 Result Declaration DateThe NEET entrance exam will be organized through computer based mode. It will be conducted on May 5, 2019. After one month of the entrance test, the result will be declared. The result will be announced on June 5, 2019. The counselling procedure will be started after the result declaration.APPLY NOW CHECK ELIGIBILITY GET UPDATESNEET Results Official ScorecardHow to Check NEET Result 2019?NTA will announce NEET 2019 Result online after answer sheets are thoroughly evaluated. The steps to check NTA NEET Result 2019 are given below:Visit the official website of NEET at National Testing Agency.Under the Online Services, click on NEET ResultA new tab will get open. Enter the required credentials to get NEET ResultsEnter your NEET credentials – Roll No. and Date of Birth and then submit it after verifying the entered details.NEET Result will be displayed on the screen.Download and take printout of NTA NEET Score CardDetails Mentioned in NEET Score CardCandidates can download NEET Score Card from the nta.ac.in . The document would be required at the time of counselling and reporting at the Institute for confirming the admission. It holds the following information:Roll NumberApplication NumberNationalityApplicant’s NameApplicant’s Date of BirthApplicant’s GenderApplicant’s Parents NameSectional-wise marks obtained in Physics, Chemistry and BiologyCategoryTotal MarksPercentile ScoreRank- All India Rank, All India Quota seat rank and State Quota rankQualifying StatusNEET Cutoff score for counselling under All India Seats 2019Sample of NEET ScorecardFormula to Calculate NEET Score and RanksWith the help of NEET Answer Key, candidates can calculate their expected scores and ranks. They can tally their answers with those given in answer key before the NEET Result is released. The marking scheme followed by NTA in NEET 2019 is given below-APPLY NOW CHECK ELIGIBILITY GET UPDATESFor every correct answer, 4 marks are rewarded.For marking wrong option, -1 is deducted as penalty.No mark is deducted for un-attempted questions and marking more than 1 option.Total Marks = 180 MCQs X 4 = 720 MarksCheck NEET Exam PatternHow is NEET Score Calculated?NEET Scores can be calculated by following the given formula-NEET Score=(Total number of correct answers x 4) – (Total number of incorrect answers x -1)NEET Rank vs Marks 2019NEET Rank is the main factor while preparing the merit list. The marks secured by candidates out of 720 determine their actual ranks in final merit list of NEET. The expected range of ranks and marks is provided in the following table-Ranks RangeMarks RangeRanks RangeMarks Range1-5681-69038,671-42,664391-4006-9671-68043,751-48,025381-39010-16661-67049,140-53,692371-38017-31651-66054886-60,006361-37032-65641-65061,286-66,854351-36066-80631-64068,197-73,907341-35081-92621-63081,674-75,426331-34092-160611-62082,464-89,970321-330172-246601-61091,617-98,710-311-320272-363591-6001,00,625-1,08,255301-310388-531581-5901,09,875-1,18,148291-300547-781571-5801,20,177-1,28,941281-290813-1,132561-5701,31,185-1,40,219271-2801,200-1,616551-5601,42,586-1,52,352261-2701,728-2,308541-5501,54,842-1,65,169251-2602,441-3,298531-5401,68,075-1,78,876241-2503,503-4,473521-530181,431-194813231-2404,708-5,972511-5201,96,386-2,10,183221-2306,257-7,696501-5102,12,003-2,25,343211-2208,032-9,570491-5002,28,954-2,42,788201-2109,958-11,594481-4902,46,509-2,61,169191-20011,993-13,926471-4802,65,271-2,80,794181-19014,358-16,342461-4702,85,115-3,01,394171-18016,795-18,977451-4603,06,153-3,23,646161-17019,548-22,114441-4503,28,574-3,46,874151-16022,756-25,447431-4403,52,020-3,71,811141-15026,179-29,211421-4303,77,662-3,98,105131-14029,973-33,388411-4204,04,017-4,28,905120-13033,794-37,770401-410--Check How to use NEET 2019 Rank Predictor?What is NEET Percentile Score?NEET Percentile Score is the minimum score which should be obtained to qualify NEET 2019.It is calculated on the basis of candidate’s relative position in comparison to other rank holders.NEET All India Rank is calculated on the basis of absolute score obtained by them in NEET.The following table represents the NEET qualifying percentile for counselling:CategoryQualifying Marks out of 720Qualifying PercentileUnreserved Category (UR)36050 PercentileScheduled Caste (SC)/Scheduled Tribes (ST)/ Other Backward Classes (OBC)28840 PercentilePhysically Handicapped (PH)324 (UR), 288 (Reserved Category)45 PercentileCandidates can calculate their NEET percentile score on the basis of rank secured by them. The exact formula for calculating NEET percentile is:NEET 2019 Answer KeyThe answer key will be released through online mode.Candidates will be able to check the OMR Sheets and Answer keys of NEET via online mode.With the help of answer key, students can calculate their rough marks they are going to score in the examination.The NEET answer key 2019 will bear the correct answers of all the questions asked in the examination.The authority will also provide the facility to challenge against the answer key. Students may challenge the answer key within the prescribed dates.Candidates will challenge the answer key via online mode with the payment of INR 1000/- per answer.If the challenge is accepted by the authority, the payment will be refunded to candidates.Steps to Check NEET Answer Key 2019Visit the official website of NEET 2019Click on the link “Download NEET 2019 Answer Key”The answer keys of all sets will appear on the screenDownload it and match your responsesSave it for further useHow to Challenge NEET Answer Key 2019?At the official website, login using your registration number and passwordClick on “Challenge NEET Answer Key” tab and submit itSelect the question you wish to challengeSelect your answer and give justified reasoningsClick on “Confirm” and pay INR 1000 as fee through Debit/ Credit cardTake a printout of the confirmation slip which contains details of the answer key challenged.NEET 2019 Merit ListNTA prepares a Merit List on the basis of directives of the Supreme Court of India, DGHS, MCI and DCI soon after the announcement of NEET Results.No category wise merit list is declared; there will be only a centralized merit list for all the NEET 2019 qualified candidates.The Merit list is prepared for seats under 15% All India Quota.The number of candidates in the merit list would be equal to the number of seats available for allotment in MBBS/BDS courses under 15% All India Quota.States will prepare their own merit list based on AIR and category of the candidate.Merit list for All India Quota is prepared on the basis of AIR and NEET Reservation Policy.There will be a waiting list equal to 4 times the number of seats available for 15% All India Quota.The merit list for NEET will be announced on its official website. Candidates are advised to check the official website as well as the websites for all NEET Participating Institutes regularly.Note: The counselling for admission to AFMC will be conducted by the Directorate General of Armed Forces Medical Service, Ministry of Defence, Government of India.Tie BreakerIn case two or more candidates have secured the same marks in NEET 2019, tiebreaker (inter-se-merit) of such candidates will be determined in order of preference as mentioned below-Higher marks secured in Biology in NEET will be preferred.If the tie persists, candidate obtaining higher marks in Chemistry in NEET will be preferred.If the tie persists, candidate with less number of wrong answers in will be taken into consideration.If the tie persists, candidate older in age will be preferred for merit ranking.What After NEET Results 2019?NTA releases NEET Cut Off along with the declaration of results. It is the minimum marks required to qualify NEET. There are separate cut off marks for different categories.The cut off is dependent on :Total number of seats available,Number of candidates appearing in the examDifficulty level of the NEET exam.NEET Cut off is released for 2 quotas– All India Quota and State Level Quota. Top 15% scorers in NEET will be allowed to participate directly in the NEET counselling process under 15% All India Quota. The expected NEET cutoff for 2019 is mentioned below:NEET Cut off under 15% All India Quota SeatsNEET Cut Off under 15% All India Quota seats is depicted in the following table:CategoryNo. of CandidatesMarks RangeGeneral14315685 - 468OBC230468- 465SC2815468- 352ST1427466 - 303UR & PH430466-214OBC & PH10213-210SC & PH75209-130ST & PH23213 - 118Category-wise NEET Overall Cut Off 2019CategoryQualifying CriteriaMarks RangeNo. of CandidatesGeneral50 Percentile685 – 145171329OBC40 Percentile678 – 118175226SC40 Percentile595 – 11847183ST40 Percentile599 – 11815710UR & PH45 Percentile474 – 131437OBC & PH40 Percentile510 – 118597SC & PH40 Percentile415 – 118143ST & PH40 Percentile339 – 11836NEET Cut Off for Top 10 Medical CollegesAs per the trends of previous years, NEET Cut Off of top 10 medical colleges is depicted in the table given below. The cut offs mentioned in the following table represents the cut off for general category-Name of College2016 Closing Cut Off Rank2015 Closing Cut Off RankGovernment Medical College, Chandigarh162158Grant Medical College and Sir JJ Group of Hospitals, Mumbai408565King George's Medical University, Lucknow506386Lady Hardinge Medical College, New Delhi263117Maulana Azad Medical College, New Delhi4428Pt. Bhagwat Dayal Sharma Post Graduate Institute of Medical Sciences, Rohtak1035674Seth G.S. Medical College, Mumbai408216Stanley Medical College, Chennai22642039University College of Medical Sciences, New Delhi12897VMMC & Safdarjung Hospital, New Delhi1065NEET 2017 ToppersGiven below is a list of all NEET Exam toppers of last year:Candidate NameAll India RankScore ObtainedNavdeep Singh1697Archit Gupta2695Manish Moolchandani3695Sankeerth Sadanand4692Abhishek Dogra5691Derik Joseph6691Kanishk Tayal7691Nikita Goyal8690Aryan Raj Singh9690Tanish Bansal10690NEET 2017 StatisticsCategoryRegistered CandidatesAppearedQualifiedMale497043473305266221Female641839616772345313Transgender885Total11388901090085611539*NEET 2019 CounsellingNEET Counselling is expected to start after July 12, 2019. Candidates who have qualified in NEET 2019 and whose names have appeared in the final merit list will be eligible for NEET CounsellingThe counselling for 15% AIQ seats and seats in Deemed and Central Universities will be conducted by MCC.The merit list for centralized counselling will be drawn according to candidates’ AIQ Rank.Candidates willing to apply for admission in State Medical Colleges/Universities/Institutions through NEET merit list should follow the instructions of the State Government and authorities of the medical/dental colleges /Institution/University concerned for the counselling process.State Merit List will be drawn on the basis of candidates’ All India Rank.Most Preferred Medical Colleges by Top 100 ScorersNEET 2019 Results FAQsQues. How will the three (3) attempts of NEET be calculated? Should it be consecutive or not?Ans. No. Information regarding the ways to calculate NEET attempts is not listed in the information brochure. The attempts will be calculated from 2019 onwards. For example, if a candidate has appeared in NEET 2019, they will be eligible for two more attempts provided they have not crossed the upper age limit.Ques. How will the merit list be prepared for 15% All India Quota seats?Ans. Based on All India Rank and the reservation roster provided by DGHS, merit list for All India quota will be prepared.Ques. How to calculate NEET Score and Rank?Ans.NEET Percentile Based on Highest MarksEvaluation of percentile on the basis of marks can easily be done. We are providing you Percentile formula which based on marks scored by the contenders.The formulae for calculating Percentile on the basis of Marks:Individual Score X 100 / Marks obtained by top scorerPercentile based on rankEvaluation of percentile on the basis of rank is somewhat difficult in comparison to the first one. Have a look at the Percentile formula which is based on rank, gained by the aspirants.The formulae for calculating Percentile on the basis of Rank:Total appearing candidates – Rank secured/ Total appearing candidates’ X 100Ques. How will the Tie-breaking problem be dissolved?Ans. In case candidates score equal marks in the NEET examination, tie-breaking method will be considered.The candidate who obtains a higher mark in Biology as per the result of NEET 2019 will be given a better rank.If Biology marks are equal, candidates securing higher marks in Chemistry as per the NEET result 2019 will be given a higher rank in NEET 2019.If both Biology and Chemistry marks as per NEET result are equal, then the candidate who has attempted less number of wrong questions in all subjects of NEET 2019 will be given better rank.If all of the above criteria are unable to prove effective in breaking tie in NEET 2019 result, then the candidate older in age will be given preference for NEET admissions.Ques. Who will conduct Counselling after NEET 2019?Ans.For 15% All India Quota seats, counseling will be conducted by the DGHS.For seats under the control of States/Universities/Institutions, their respective authorities will conduct the counseling.Ques. What is the reservation criteria for 15% all India quota seats?Ans. In 15% all India quota seats, 15% seats are reserved for SC candidates, 7.5% for ST and 27% for OBC category candidates.Ques. What is the procedure of online seat allotment?Ans. The process steps are given here -Main Counselling Registration.Exercising of Choices, indicative seats and Locking of choices.Process of Seat Allotment – Round 1.Round 1 Result Publication.Reporting at the allotted Medical/Dental College against 1st Round.Publication of net vacant seats.Fresh Choice Submission by eligible candidates for Round 2.Process of Seat Allotment – Round 2.Round 2 Result Publication.Reporting at the Medical / Dental Colleges /Institutions against Round 2.END of 15 % All India Counselling and reverting of Non Joined, Not allotted seats to state quota.Ques. Will there be negative marking in NEET 2019?Ans. Yes. Each correct answer will get 04 marks and for each incorrect response, 01 marks will be deducted from the total score.Ques. How long is the result of NEET 2019 valid?Ans. The result of successful candidates of NEET 2019 is valid for the academic session 2019 – 20 only.Ques. Can one see images of OMR sheets and Responses on website?Ans. The images of OMR Sheet and Responses captured by the machine will be displayed on the website National Testing Agency. For exact date of hosting, candidates may remain in touch with website.Ques. I did not have a graduation degree, However, I have completed a professional degree, Can I apply for NEET?Ans. No, only candidates who are graduate from a recognized University can apply for NEET. No, professional degree is accepted as a graduation degree.Ques. Is there any provision to challenge the response(s) against the OMR grading?Ans.Candidates will be given an opportunity to submit representation against the OMR grading on payment of INR 1000 per response challenged.Rs.1000/- per response will be refunded to candidate if his/her challenge is accepted by the Board.Ques. How many marks should we obtain to qualify for NEET exam?Ans. Here we are giving the qualifying marks for NEET 2019 examination:CategoryQualifying Marks (Out of 720)General (UR)360OBC/SC/ST288General (UR) – PH324OBC/SC/ST – PH288*The article might have information for the previous academic years, which will be updated soon subject to the notification issued by the University/CollegeGET MORE INFOFREE CounsellingSAMPLE PAPERSupcoming examsIUET March 11, 2019RUET April 22, 2019GCET May 8, 2019IPU-CET May 13, 2019AIIMS-MBBS May 25, 2019top coursesMBBSBDSBE/B.TECHB.SCB.COMOther Colleges In The Same GroupJaipur National University - [JNU] Jaipur, RajasthanSardar Patel University - [SPU] Vallabh Vidyanagar, GujaratKalinga Institute of Industrial Technology - [KIIT] Bhubaneswar, OrissaVeer Narmad South Gujarat University - [VNSGU] Surat, GujaratT.D. Medical College - [TDMC] Alappuzha, Kerala

People are often confused about literary agents—how they work, what they do, and how to find one. The time-honored saying is that it’s easier to find an editor than an agent, and to some extent, that’s still true. But in fact people go about looking for agents in a way that is often ineffective and, to be honest, all wrong.The first thing to understand is how agents work. The best comparison is to compare literary agents to real estate agents, or realtors. If you own a property (we’ll say a home) and want to list it for sale, it’s usually no problem to find a listing agent. But finding the wrong realtor will result in frustration or even loss when the deal is finally made or even if it is made at all. All realtors are not the same. Some specialize in luxury homes, and some only deal with rural homes with property surrounding them. Some deal exclusively with industrial or commercial properties only, and some only with vast swaths of empty land that might be for sale. Some, of course, will take on any listing of any or all of these, and they will even represent shacks and derelict houses; some will represent mobile homes or temporary housing; some will deal only with condos and apartments, and some will take on anything they think they can sell. Even within large realty companies, all of these may exist, all working side-by-side, so if you call a large agency in search of a realtor, you’ll be routed to one who usually takes on property such as you own.That makes sense. If you have a modest home, being represented by a realtor who only services clients with multi-million dollar mansions will probably do you no good; similarly, if you have an estate worth millions, you don’t want to list it with a realtor whose client list consists of people with only a few thousand to spend.Literary agents work the same way. Some represent only one kind of book. Some specialize in nonfiction, some in fiction. Some specialize in various categories of both of these, say autobiography and memoir, or political and economic writing, scientific writing, sociological commentary, biography and tell-all works, or other categories of nonfiction. Some only want to represent what are sometimes called “coffee-table books,” or oversized and heavily illustrated volumes. In fiction, there are agents who specialize in romance novels, historical fiction, detective and crime, science fiction and horror, military fiction, spy fiction, western fiction, what is sometimes called “literary fiction,” children’s and Young Adult fiction, and so forth. This is because a literary agent’s client base is directed toward the editors and publishing houses that he or she works with most ordinarily.Literary agents usually sell the works they represent through personal contact, either face-to-face or on the phone; there are only so many hours in a day, so many days in a week, so they tend to narrow their editorial contacts to people who work for publishing houses or imprints in a specific category of interest. They may take on multiple categories, and certainly there are agents who will take anything if they think they can sell it. But the key for the writer, just as the key for the homeowner is to locate an agent who has the most sharply honed interest in the kind of property that is being offered for sale. They are more in touch with the market in that category, have the greatest number of potential buyers on their list, and, not incidentally, have spent a great deal of their time and interest in studying the market in a given category, so they know what sells, what won’t.This brings us to money. No writer should ever, ever, ever, ever, ever, ever pay a literary agent to represent a work. Just as no homeowner wanting to sell would pay a realtor for a listing, no writer should pay an agent for representation. The agent makes money when the book is bought by a publisher. (The going rate is 15%, but there are both higher and lower rates sometimes offered or demanded. This is negotiable, in other words). It works roughly the same way as in real estate. A realtor makes a commission on a sale—how much is negotiable at the time of the listing and varies widely in real estate. As a rule writers are asked to sign contracts with the agent, usually offering exclusivity to the agent for the work, and often asking the writer to agree to terms regarding any royalties or residuals that may come from the work once it’s been published. These contracts have limits, and they are negotiable. A realtor doesn’t make any money from residuals or future earnings from a property once it’s sold; however, they do have exclusivity as the listing agent for the property, and they will make money if it’s sold, no matter who actually sells it, even the owner, so long as the listing is active. Listings have time limits, too. But no sensible home owner wanting to sell would ever pay a realtor to take a listing. That’s just stupid.Importantly, though, agents do not make a dime until the work is sold. There are numerous literary agents who do demand a reading fee, a representation fee, or require other payment for other reasons from the writer. Only a fool would pay this, since there is no guarantee of sale. Unscrupulous agents often make their entire income from these fees and may or may not represent any book they take for representation. There is no way of knowing whether they are marketing the book or not. Thee are so many ways a writer can be duped about that that it’s a waste of time listing them. Just remember the rule, never, never, never, never, never pay up front for representation.Another similarity between literary agents and realtors arises here. If a realtor deals only in one kind of property, say luxury homes, town homes, or condos, the properties listed are all offered to a potential buyer. A particular home that is less than saleable might be used as a hook to get a buyer interested before the realtor shifts the buyer’s interest to another property the realtor has listed, instead. The opposite can happen, as well. Realtors work for the seller, technically, but they make money from the buyer. So it’s in their interest to “up sell” a buyer, moving interest from one modestly priced property to a more expensive one. In the same way, literary agents work with a limited number of editors whom they know. They may arrange a meeting to submit a book for purchase to a given editor who they know publishes only one or two categories of books, and use that as a “hook” to get the editor’s attention, but then shift attention to another or different title that the agent thinks will make more money. Sometimes, these can even be packaged. Additionally, if an editor expresses distaste for a particular submission, one more attractive may be offered instead. There’s nothing unscrupulous about this. But the writers have no way of knowing how much enthusiasm the agent is putting behind the book offering to an editor. More to the point, if a previous deal with an editor has gone sour for some reason, the whole marketing plan may be geared toward not compounding the offense.Here a major difference in the analogy arises, though. There is no limit to the number of buyers for real estate. There is a limit to the number of editors who buy books, particularly in certain categories of writing. Forty years ago, agents might have a hundred or more editors to choose from for approach; today, it’s more like a few dozen. So the market has shrunk. Real estate never shrinks in terms of those wanting to buy it. But then real estate for sale doesn’t shrink, either. Neither, as it happens, do literary properties. There are more people trying to sell books today than ever before, and there are fewer people publishing books today than there have been in the past 150 years.Literary agents make their money, then, from sales. This is why they are reluctant or absolutely refuse to market books to small presses and university presses or certainly to self-publishing outfits. There’s no advance against royalty there, and the print runs are so small that there’s little likelihood that there will be any sales to speak of. So they won’t make anything, and the reward is not worth the effort. There are some—perhaps many—exceptions to this rule, but in that case, it’s usually because the agent sees some value in the book’s potential, possibly for film rights, translation rights, or mass market reprint rights, all of which could provide a payday later on.Another misconception about agents is that their principal job is to sell the book. That actually is only a small part of their obligation. They also can help the writer read and understand a publishing contract, protect the writer from clauses, sometimes hidden in the boiler plate of a contract, that might wind up costing money later on. The also negotiate contracts for residual and after-market sales, translations, film rights, audio rights, electronic reprint rights, paperback rights, and so forth. Protecting the writer’s rights to a literary property is as important as selling it in the first place. It’s in the agent’s interest, also, as there are commissions to be made from the sale of any of these after the initial publication. They also can be useful in securing the return of rights to the author after the initial publishing lifetime of the book is over, and there’s the matter of obtaining and reading royalty statements and making sure that all rights are secure. This is a contractual arrangement, and writers are encouraged to read their agency contracts with a close eye, as well.So the last point has to do with how a writer locates an agent. That’s hard, no two ways about it, but it’s a whole lot easier than trying to find an editor without one. These days, almost no legitimate publishing house accepts unsolicited and/or unagented submissions. Literary agencies have become virtual clearing houses for publishers, filtering out the worthless and advancing work that has some promise.Unlike realtors (another exception) there is no regulatory agency governing literary agencies. Some states—California, for example—require licenses for film agencies, but most states don’t. All that’s required to become a literary agent is to declare that one is such. There are individual agents operating solo, agents who establish their own agencies with sub-agents or associates, large literary agencies that employ dozens or even hundreds of agents on their staff, and all manner of agents in between. There are even agents who have nothing more than a website and maybe a post office box, sometimes. Some of these are marginally criminal. Each of the legitimate and reliable agents, either collectively or individually, though, tends to specialize, as a rule, in one or more categories of writing. They might take on a book outside those categories, but it’s not likely they will. So matching the kind of book one has to the kind of agent one needs is a significant part of the process.A writer would be well-advised to do a good deal of homework in the matter before approaching an agent. Publications such as Writer’s Market and Literary Market Place offer more or less comprehensive lists of legitimate agents and agencies, with the particulars of what they want, their fees, if any, and requirements as well as recommendations for the best way for a writer to approach them. It’s worthwhile to study these lists, as well as to look up agents’ websites and blogs, if they have such, to see what they want, how they want it. On-line submission sites such as Submittable.com. can be remarkably useful, as well. Blindly picking an agent only by name might have worked thirty-five or forty years ago; it won’t do much for a writer today other than to take up a lot of time and waste energy.Writer’s Market’s recommendations for how to approach agents are worth looking at, but not to follow absolutely. My experience has been that a short, to the point, cover letter over a number of sample pages or chapters is the best approach. Sending in the entire book can also work, unless the agency stipulates that such absolutely should not be done. Today, almost all agents want electronic submission; that’s unavoidable, and it’s also unfortunate, as it’s hard to read effectively on screen, and it’s easy to hit the “delete” key. Perhaps too easy. Having to unwrap and then handle a sheaf of pages, then, if a rejection is warranted, to have to pack it back up and put it in a provided SASE, required an agent at least to look at it, read more than the title, and then decide. But that’s the way of the world, just as property buyers review properties for sale via on-line websites, slide shows or videos, these days, rather than make a series of appointments for on-site viewing.Regardless, it’s a foregone conclusion that these days it’s almost impossible to gain a quality publication without a literary agent, and the relationship a writer has with the agent if vitally important. Some agents will act as “editors” in the sense that they may like a book but want it revised or edited before representation. They might ask for cuts or additions, corrections or changes. This may be a requirement for representation, and the decision about whether or not the work should be done rests entirely with the writer. Agent might also recommend “book doctors” or editors to work on a book they think has promise. That is also a decision a writer must make, and a particular aspect of that has to be whether or not there’s an assurance of acceptance for representation if the work is done, and paid for, by the writer. This assurance is seldom forthcoming; but if it is offered, the writer should get that promise in writing, with some form of guarantee of refund if it doesn’t happen.Just as there are unscrupulous realtors who will demand that a home be remodeled or repaired or otherwise altered and then will recommend a contractor who will do the work, then pay them a kick-back, unscrupulous literary agents sometimes recommend specific “book doctors” who charge writers high fees for the work, then pay a kickback to the agents, who have no intention of taking the book no matter what is done to it. Writers who feel they need outside editorial help are best advised to arrange for their own “book doctors,” unless iron-clad guarantees are in place.So a writer is advised to do a good deal of homework and then to approach agencies that are most likely to accept work that they think they can sell. Importantly, the writer should be aware that an agent might reject a book even though they admire it simply because they have no idea of how they might sell it or to whom. A rejection is not a comment on a book’s quality. It’s merely an indication of whether or not an agent thinks that there is enough money in it to make it worth the time it will take to market it effectively. There is also no prohibition about approaching more than one agent at a time. Some writers might submit their work to a dozen or more agents at once, just to see if one of them “bites.” If that happens and if an arrangement is made, then the other agents should be notified immediately upon signing the deal.I don’t know if this answers your question, but it might give you perspective on literary agencies. Some writers change agents often, year to year, or book to book. One writer I know says, “The largest club in the country is made up of agents I’ve fired.” Others, myself included, have been lucky enough to remain with the same agent for years, as a trust has grown up between us. I might not always agree with him, but I respect his judgement; and I know that if he refuses a book I’ve offered it’s because he doesn’t think he can sell it, not necessarily because he doesn’t like it.

This is a great question because I don’t think most papers do a good job explaining where it comes from.Here’s the short answer: quite often we want to use samples from the policy to learn, but the sampling introduces an invalid bias to the estimation of the gradient. Consequently, we use importance sampling to correct the distribution, and once you use importance sampling with the gradient, you can simplify the result into the gradient of the log probability.Here’s the long answer.Okay, so that short answer has probably left you scratching your head about all the jumps. Lets break it down.First, lets identify the following: what is the objective function we want to maximize? From “Policy Gradient Methods for Reinforcement Learning with Function Approximation” [1] we have the following objective:[math]J(\pi) = \sum_s d^\pi(s) \sum_a \pi(s, a) Q^\pi(s, a)[/math]where the [math]d^\pi[/math] describes the (discounted) probability of reaching each state under the policy. Recall that Q(s, a) is the function that tells us what the expected reward is if we take action a from state s and then follow our policy.Although there are a lot math symbols here the idea is pretty simple: the objective is to adjust our policy so that it maximizes the probability that it will select an action with the highest possible future rewards in each state. That should match your expectation of what the objective should be: of course we want to change our policy to maximize the future rewards!If we are using neural nets and want to optimize an objective with stochastic gradients descent (or ascent in the case of an objective to maximize), then we want to be able differentiate the objective and move along the gradient.What that paper goes on to show is that the gradient of the objective has a simple form:[math]\nabla_\theta J(\pi) = \sum_s d^\pi(s) \sum_a \frac{{\partial \pi(s, a)}}{{\partial \theta} } Q^\pi(s, a)[/math]The only thing we need to differentiate is the policy, the rest of the equation stays the same.*Although that gradient looks simple there is one thorn in our side: for each state we have to sum over all actions. Why is that a thorn? Because in practice we want to use samples from our policy, but if we are using sample from our policy, we can’t rewind the clock and choose a different action in each state. Consequently, for each state, we’d be missing the contribution of all other action choices the policy might have made.This is where importance sampling comes in. If you are not familiar with importance sampling I’ll direct you to my answer explaining it [2]. The gist of importance sampling is it’s a modification for computing an expected value for some distribution when sampling from a different distribution.Here we’re using it slightly differently: we don’t really have expected value over some distribution: we just have function that is a sum over actions. But sums are a lot like uniform distributions so the same tricks of importance sampling apply.I’ll walk through the derivation here. Lets ignore all parts of the gradient except the last inner sum. We can remark on a simple identity:[math]\sum_a \frac{{\partial \pi(s, a)}}{{\partial \theta} } Q^\pi(s, a) = \sum_a \frac{\pi(s, a)}{\pi(s, a)} \frac{{\partial \pi(s, a)}}{{\partial \theta} } Q^\pi(s, a)[/math]This should make sense because all we did is multiply by the policy probability and then immediately divide by it to cancel it. Consequently, we haven’t changed anything. But now we note that this looks like an expected value where the distribution is the policy distribution:[math]\sum_a \frac{\pi(s, a)}{\pi(s, a)} \frac{{\partial \pi(s, a)}}{{\partial \theta} } Q^\pi(s, a) = E_{a\sim\pi(s, \cdot)} \left[ \frac{1}{\pi(s, a)} \frac{{\partial \pi(s, a)}}{{\partial \theta} } Q^\pi(s, a) \right][/math]What that tells us is if we have a sample of the policy, we can modify our gradient by just dividing by the probability of sampling that action. And if we have enough samples to compute an average, we’ll recover the original gradient!Okay, we’re almost there now. It’s just a bit of algebra at this point. Let remark on one more simple identity of the derivative of logs:[math]\frac{{\partial \log(f(x))}}{{\partial x}} = \frac{1}{f(x)}\frac{\partial f(x)}{{\partial x}}[/math]As you will note, the right hand side is the same form we have in our gradient, so we can just substitute the gradient of the policy divided by the policy probability with the gradient of log of the policy.[math]E_{a\sim\pi(s, \cdot)} \left[ \frac{1}{\pi(s, a)} \frac{{\partial \pi(s, a)}}{{\partial \theta} } Q^\pi(s, a) \right] = E_{a\sim\pi(s, \cdot)} \left[ \frac{{\partial \log(\pi(s, a))}}{{\partial \theta} } Q^\pi(s, a) \right][/math]And there is it! The log has appeared! Taking the gradient of the log probability tends to be more numerically stable than computing the actual density and dividing even though mathematically they’re the same thing, so we tend to prefer to keep things in the log representation.So the key here is that the log appeared *because* we introduced importance sampling so that we could sample actions instead of having to sum over actions we didn’t actually take.Afifatul Mukaroh regarding your follow up questions, note that the cross entropy function for a single “label” is the same as the negative log probability of the categorical distribution selecting that label . Consequently, you’ll often see people use the “cross entropy function” to compute the negative log probability [3].Similarly, minimizing the L2 norm of a prediction with a continuous label is the same as maximizing the log likelihood of a Gaussian distribution with a standard deviation of 1. So when actions are continuous and we have a Gaussian policy (with standard deviation of 1), people often just use the L2 norm function in code.That is, the cross entropy and L2 norm are still accomplishing the goal of maximizing the log probabiltiy for those respective classes of probability distributions.*It might seem obvious that that would be the gradient, but because the true Q-function involves environment transition dynamics, this fact is not immediately clear and requires some proof, which the paper provides.https://homes.cs.washington.edu/~todorov/courses/amath579/reading/PolicyGradient.pdfWhy doesn’t DQN use importance sampling? Don't we always use this method to correct the sampling error produced by the off-policy?Cross entropy - Wikipedia