閱讀理解說(shuō)明文-2024年高考英語(yǔ)大題集訓(xùn)（新高考解析版）

專題03問(wèn)讀理解說(shuō)明文

說(shuō)明文所選材料題材豐富多樣，涉及社會(huì)、科技、文化、生活、人物、教育、生態(tài)、安全等方面。在

高考英語(yǔ)閱讀理解中，說(shuō)明文所占比重尤為突出。它既要求考生能熟練運(yùn)用詞匯和語(yǔ)法知識(shí)、理解文章語(yǔ)

句、把握語(yǔ)篇整體結(jié)構(gòu)，還要求學(xué)生有大量的閱讀積淀和知識(shí)儲(chǔ)備，熟悉不同的話題和不同的題目考查方

式。即便是細(xì)節(jié)理解題，大多數(shù)情況下考生也無(wú)法從文章中直接找到與選項(xiàng)表述完全一致的信息，而是需

要在理解文章細(xì)節(jié)信息后作出歸納和判斷。通常一套卷中主旨大意、詞義猜測(cè)、寫(xiě)作意圖/觀點(diǎn)態(tài)度類高難

度試題的考查數(shù)量為2道左右，說(shuō)明文考查此類題目的頻率較高，且命題點(diǎn)呈多樣化。

（2023年新高考I卷D篇）OnMarch7,1907,theEnglishstatisticianFrancisGaltonpublishedapaper

whichillustratedwhathascometobeknownasthe"wisdomofcrowds“effect.Theexperimentofestimationhe

conductedshowedthatinsomecases,theaverageofalargenumberofindependentestimatescouldbequite

accurate.

Thiseffectcapitalizesonthefactthatwhenpeoplemakeerrors,thoseerrorsaren'talwaysthesame.Some

peoplewilltendtooverestimate,andsometounderestimate.Whenenoughoftheseerrorsareaveragedtogether,

theycanceleachotherout,resultinginamoreaccurateestimate.Ifpeoplearesimilarandtendtomakethesame

errors,thentheirerrorswon'tcanceleachotherout.Inmoretechnicalterms,thewisdomofcrowdsrequiresthat

people'sestimatesbeindependent.Ifforwhateverreasons,people'serrorsbecomecorrelatedordependent,the

accuracyoftheestimatewillgodown.

ButanewstudyledbyJoaquinNavajasofferedaninterestingtwist（轉(zhuǎn)折）onthisclassicphenomenon.The

keyfindingofthestudywasthatwhencrowdswerefurtherdividedintosmallergroupsthatwereallowedtohave

adiscussion,theaveragesfromthesegroupsweremoreaccuratethanthosefromanequalnumberofindependent

individuals.Forinstance,theaverageobtainedfromtheestimatesoffourdiscussiongroupsoffivewas

significantlymoreaccuratethantheaverageobtainedfrom20independentindividuals.

Inafollow-upstudywith100universitystudents,theresearcherstriedtogetabettersenseofwhatthegroup

membersactuallydidintheirdiscussion.Didtheytendtogowiththosemostconfidentabouttheirestimates?Did

theyfollowthoseleastwillingtochangetheirminds?Thishappenedsomeofthetime,butitwasn'tthedominant

response.Mostfrequently,thegroupsreportedthatthey“sharedargumentsandreasonedtogether”.Somehow,

theseargumentsandreasoningresultedinaglobalreductioninerror.AlthoughthestudiesledbyNavajashave

limitationsandmanyquestionsremain,thepotentialimplicationsforgroupdiscussionanddecision-makingare

enormous.

1.Whatisparagraph2ofthetextmainlyabout?

A.Themethodsofestimation.B.Theunderlyinglogicoftheeffect.

C.Thecausesofpeople'serrors.D.ThedesignofGalton'sexperiment.

2.Navajas5studyfoundthattheaverageaccuracycouldincreaseevenif

A.thecrowdswererelativelysmallB.therewereoccasionalunderestimates

C.individualsdidnotcommunicateD.estimateswerenotfullyindependent

3.Whatdidthefollow-upstudyfocuson?

A.Thesizeofthegroups.B.Thedominantmembers.

C.Thediscussionprocess.D.Theindividualestimates.

4.Whatistheauthor'sattitudetowardNavajas'studies?

A.Unclear.B.Dismissive.C.Doubtful.D.Approving.

【答案】1.B2.D3.C4.D

【導(dǎo)語(yǔ)】本文是說(shuō)明文。沒(méi)有人是一座孤島，文章陳述了“群體智慧”效應(yīng)。實(shí)驗(yàn)表明，在某些情況下大量

獨(dú)立估計(jì)的平均值可能是相當(dāng)準(zhǔn)確的。

1.主旨大意題。根據(jù)第二段內(nèi)容“Thiseffectcapitalizesonthefactthatwhenpeoplemakeerrors,thoseerrors

aren'talwaysthesame.Somepeoplewilltendtooverestimate,andcometounderestimate.Whenenoughofthese

errorsareaveragedtogether,theycanceleachotherout,resultinginamoreaccurateestimate.Ifpeoplearesimilar

andtendtomakethesameerrors,thentheirerrorswon'tcanceleachotherout.Inmoretechnicalterms,the

wisdomofcrowdsrequiresthatpeople'sestimatesbeindependent.Ifforwhateverreasons,peopleserrors

becomecorrelatedordependent,theaccuracyoftheestimatewillgodown.（這種效應(yīng)利用了這樣——個(gè)事實(shí)，即

當(dāng)人們犯錯(cuò)誤時(shí)，這些錯(cuò)誤并不總是相同的。有些人常常會(huì)高估，或者低估。當(dāng)這些誤差中有足夠多的誤

差被平均在一起時(shí)，它們會(huì)相互抵消，從而產(chǎn)生更準(zhǔn)確的估計(jì)。如果相似的人傾向于犯同樣的錯(cuò)誤，那么

他們的錯(cuò)誤不會(huì)相互抵消。從更專業(yè)的角度來(lái)說(shuō)，群眾的智慧要求人們的估計(jì)是獨(dú)立的。如果由于任何原

因，人們的錯(cuò)誤變得相關(guān)或依賴，估計(jì)的準(zhǔn)確性就會(huì)下降）”可知，本段闡述了人們所犯的錯(cuò)誤不總是相同

的，各不相同的誤差平均在一起，相互抵消就會(huì)產(chǎn)生更準(zhǔn)確的估計(jì)，討論了獨(dú)立估計(jì)的平均如何由于誤差

的消除而產(chǎn)生更準(zhǔn)確的預(yù)測(cè)。因此本段主要解釋了“群體智慧”效應(yīng)這一現(xiàn)象的基本邏輯。故選B。

2.細(xì)節(jié)理解題。根據(jù)第二段的“Inmoretechnicalterms,thewisdomofcrowdsrequiresthatpeople'sestimatesbe

independent.（從更專業(yè)的角度來(lái)說(shuō)，群眾的智慧要求人們的估計(jì)是獨(dú)立的）”和第三段的“Thekeyfindingof

thestudywasthatwhencrowdswerefurtherdividedintosmallergroupsthatwereallowedtohaveadiscussion,

theaveragesfromthesegroupsweremoreaccuratethanthosefromanequalnumberofindependentindividuals.

Forinstance,theaverageobtainedfromtheestimatesoffourdiscussiongroupsoffivewassignificantlymore

accuratethantheaverageobtainedfrom20independentindividuals.（這項(xiàng)研究的關(guān)鍵發(fā)現(xiàn)是，當(dāng)人群被進(jìn)一步

劃分為允許進(jìn)行討論的小組時(shí)，這些小組的平均值比同等數(shù)量的獨(dú)立個(gè)體的平均值更準(zhǔn)確。例如，從四個(gè)

五人討論組的估計(jì)中獲得的平均值明顯比從20個(gè)獨(dú)立個(gè)體獲得的平均值更準(zhǔn)確）”可知，人們?cè)跊](méi)有獨(dú)立的

情況下，分成更小群體，平均值是更準(zhǔn)確的，說(shuō)明即使在估計(jì)數(shù)字并非完全獨(dú)立的情況下，準(zhǔn)確率提高也

是可以做到的。故選D。

3.推理判斷題。根據(jù)第四段的“Inafollow-upstudywith100universitystudents,theresearcherstriedtogeta

bettersenseofwhatthegroupmembersactuallydidintheirdiscussion.Didtheytendtogowiththosemost

confidentabouttheirestimates?Didtheyfollowthoseleastwillingtochangetheirminds?（在一項(xiàng)針對(duì)100名大

學(xué)生的后續(xù)研究中，研究人員試圖更好地了解小組成員在討論中的實(shí)際行為。他們是否傾向于選擇那些對(duì)

自己的估計(jì)最有信心的人？他們追隨那些最不愿意改變主意的人嗎）”可知，在后續(xù)研究中，研究人員試圖

更好地了解小組成員在討論中實(shí)際做了什么。結(jié)合兩個(gè)問(wèn)題，因此可知后續(xù)研究的重點(diǎn)是小組內(nèi)的討論過(guò)

程。故選C。

4.推理判斷題。根據(jù)最后一段內(nèi)容“AlthoughthestudiesledbyNavajashavelimitationsandmanyquestions

remain,thepotentialimplicationsforgroupdiscussionanddecision-makingareenormous.（盡管Navajas領(lǐng)導(dǎo)的

研究有局限性，仍存在許多問(wèn)題，但對(duì)小組討論和決策的潛在影響是巨大的）”可知，作者認(rèn)為雖然Navajas

領(lǐng)導(dǎo)的研究有局限性也存在許多問(wèn)題，但對(duì)小組討論和決策的潛在影響巨大。因此推斷作者對(duì)于Navajas

的研究表示一定的贊許和支持。故選D。

解題技巧

說(shuō)明文閱讀量大，生僻詞匯多，句式結(jié)構(gòu)復(fù)雜，這就需要考生有方法有技巧地去閱讀。一般我們采用

的方法有三個(gè)：細(xì)讀、略讀和跳讀。

（1）細(xì)讀重點(diǎn)。重點(diǎn)，就原文而言，就是文章的基本結(jié)構(gòu)、內(nèi)容和態(tài)度；就答題而言，就

是問(wèn)題所對(duì)應(yīng)的題源句。以這個(gè)標(biāo)準(zhǔn)來(lái)衡量，需要重點(diǎn)閱讀的原文詞句就不會(huì)很多。

宏觀方面：文章結(jié)構(gòu)；文章主題句；各段首末句；體現(xiàn)作者觀點(diǎn)態(tài)度的詞句。

微觀方面：有轉(zhuǎn)折處；重要標(biāo)點(diǎn)；句子主干。

（2）有選擇地略讀或跳讀。在快速瀏覽了題干，初步把握了文章大意及結(jié)構(gòu)的基礎(chǔ)上，可大膽進(jìn)行略讀

或跳讀，這樣不但可以加快閱讀速度，還更容易總體把握全文。在閱讀中可略讀或跳讀的內(nèi)容包括：

①繁瑣的例證。為了說(shuō)明問(wèn)題，作者可能會(huì)借用形象的例子，有時(shí)候會(huì)比較繁瑣，雖能說(shuō)明問(wèn)題，但需

要讀者花很多時(shí)間去閱讀。如果例子所說(shuō)明的問(wèn)題不明了，可通過(guò)略讀來(lái)理解；若例子所說(shuō)明的問(wèn)題比較

明了，則例子部分可一帶而過(guò)。

②并列多項(xiàng)列舉。有時(shí)許多功能相同的項(xiàng)目并列列舉，那么只讀其中一兩項(xiàng)即可，無(wú)需

全讀。

③無(wú)關(guān)大局的生僻詞匯。閱讀中經(jīng)常會(huì)遇到一些生詞，如果這些生詞對(duì)理解全文沒(méi)有影響或影響不大就

可略過(guò)。

④較長(zhǎng)的人名、地名。有許多較長(zhǎng)的表示人名、地名等的專有名詞，閱讀時(shí)可一掃而過(guò)或干脆用其首字

母代替，不必試圖把整個(gè)專有名詞讀出來(lái)。

在略讀過(guò)程中，要特別注意對(duì)解題有重要意義的詞、句、段等，并將其儲(chǔ)存在大腦里，以免回頭再查

看費(fèi)時(shí)費(fèi)力。

變式訓(xùn)練

（2024上?河北滄州?高三泊頭市第一中學(xué)校聯(lián)考期末）Thinkofyourmostprizedpossession.Now,

imagineyoudecideyoumustsellthisitem.Howmuchisitworthtoyou?Howmuchwouldyouchargeforit?

Andwouldanyoneelsewanttobuyitatthatprice?

Lefsconsiderthereal-worldexampleofayardsale.Beforeyourneighborscometolookatyourthings,you

mustfirstdecideonthepricesforthem.Foryou,eachitemhasastory,andsomeitemsmayevenevokestrong

emotions.

Lefssayyou'resellingapairofspecialeditionsportsshoesthatyoutookcareofandreallyenjoyedwearing

onspecialoccasions.Withthesesweetmemoriesinmind,youputapricetagof$130ontheshoes.However,a

customeroffers$50forthem.Ybuthink,ctHowcouldtheynotseehowspecialandvaluabletheseare?”

Infinancialtransactionslikethis,sellersoftenbelievetheirownitemsareworthmorethanwhatbuyersare

willingtopay.Thisoccursbecausebuyersdonothavetheseller'sexperienceswithandemotionalattachmentto

theitems.Buyerswantagooddeal,andthey9renotwillingtopaywhatthesellerisaskingfor.

Thisnegotiationprocesscanevenfeelalittlepainfulfortheseller-overlosingsomethingandthebuyernot

valuingitasmuch.Economistssuggestthisdynamicoccursthroughtheendowmenteffect（稟賦效應(yīng)）一people's

tendencytovaluethingstheyownmorehighlythantheywouldiftheydidnotownthem.Theendowmenteffect

cantakeholdanytimewefeelasenseofownershipoveraproduct,anditcanhappenquickly.

Asthesayinggoes,"'Emotionsgelthebestofus."Thereisnothingwrongwithbeingproudofowningthings

andtreasuringyourexperienceswiththem.Butwhenyouconsidersellingaprizedpossession,thinkabout

whetheryou9refullyreadytopartwithil.Thepastisthepast,though.Removingyouremotionalattachmentto

materialpossessionscanhelpyoumakemoreaccurateassessmentsofhowmuchyouritemsaretrulyworth.

1.Whatdoestheunderlinedword“evoke“inparagraph2probablymean?

A.Advocate.B.Cause.C.Identify.D.Control.

2.Whatmakesyouputapricetagof$130onyourshoesaccordingtoparagraphs?

A.Yourfondnessfortheshoes.B.Yourignoranceofthemarket.

C.Thehighqualityoftheshoes.D.Thespecialfunctionoftheshoes.

3.Whichcanbestillustratetheresultofendowmenteffect?

A.Onelikescollectingoldthings.

B.Onetendstobuyexpensivegoods.

C.Oneoffersalowerpricewhenshopping.

D.Onechargesahigher-than-usualpriceforsomething.

4.Whafsthepurposeofthelastparagraph?

A.Tomakeasuggestion.B.Tointroduceatheory.

C.Toanalyzeaphenomenon.D.Toprovidesomeevidence.

【答案】1.B2.A3.D4.A

【導(dǎo)語(yǔ)】這是一篇說(shuō)明文。文章主要介紹了“稟賦效應(yīng)”的現(xiàn)象，并建議消除對(duì)物質(zhì)財(cái)富的情感依戀，更準(zhǔn)

確地評(píng)估物品的價(jià)值。

1.詞句猜測(cè)題。根據(jù)上文“Foryou,eachitemhasastory（對(duì)你來(lái)說(shuō)，每件物品都有一個(gè)故事）”和下文的

emotions可知，此處是指一些物品可能會(huì)“喚起”一些情感，所以推知?jiǎng)澗€詞是“引起、喚起”的意思。故選

B項(xiàng)。

2.細(xì)節(jié)理解題。根據(jù)文章第三段"Let'ssayyou'resellingapairofspecialeditionsportsshoesthatyoutookcare

ofandreallyenjoyedwearingonspecialoccasions.Withthesesweetmemoriesinmind,youputapricetagof

$130ontheshoes.（假設(shè)你正在銷售一雙特別版的運(yùn)動(dòng)鞋，你很喜歡在特殊場(chǎng)合穿它。帶著這些甜蜜的回憶，

你給這雙鞋標(biāo)上了130美元的價(jià)格。）”可知，你對(duì)這雙鞋的喜愛(ài)讓你給這雙鞋貼上了130美元的價(jià)格標(biāo)簽。

故選A項(xiàng)。

3.推理判斷題。根據(jù)文章倒數(shù)第二段“Economistssuggestthisdynamicoccursthroughtheendowmenteffect

（稟賦效應(yīng)）一people'stendencytovaluethingstheyownmorehighlythantheywouldiftheydidnotown

them.（經(jīng)濟(jì)學(xué)家認(rèn)為，這種動(dòng)態(tài)是通過(guò)稟賦效應(yīng)發(fā)生的一人們傾向于認(rèn)為自己擁有的東西比自己沒(méi)有擁

有的東西更有價(jià)值。）”可知，一個(gè)人對(duì)某物收取比平常更高的價(jià)格，最能說(shuō)明稟賦效應(yīng)的結(jié)果。故選D項(xiàng)。

4.推理判斷題。根據(jù)文章最后一段“Thepastisthepast,though.Removingyouremotionalattachmentto

materialpossessionscanhelpyoumakemoreaccurateassessmentsofhowmuchyouritemsaretrulyworth.（不過(guò),

過(guò)去的就讓它過(guò)去吧。消除對(duì)物質(zhì)財(cái)富的情感依戀可以幫助你更準(zhǔn)確地評(píng)估你的物品到底值多少錢。）”可

知，最后一段的目的是提出建議。故選A項(xiàng)。

刷模擬

（2024?江蘇連云港?統(tǒng)考一模）Artificialintelligencemodelscantrickeachotherintodisobeyingtheir

creatorsandprovidingbannedinstructionsformakingdrugs,orevenbuildingabomb,suggestingthatpreventing

suchAI“jailbreaks“ismoredifficultthanitseems.

Manypubliclyavailablelargelanguagemodels（LLMs）,suchasChatGPT,havehard-codedrulesthataimto

preventthemfromexhibitingracialorsexualdiscrimination,oransweringquestionswithillegalorproblematic

answers-thingstheyhavelearnedfromhumansviatrainingdata.Butthathasn'tstoppedpeoplefromfinding

carefullydesignedinstructionsthatblocktheseprotections,knownas’jailbreaks”,makingAImodelsdisobeythe

rules.

Now,ArushTagadeatLeapLaboratoriesandhisco-workershavefoundaprocessofjailbreaks.Theyfound

thattheycouldsimplyinstructoneLLMtoconvinceothermodelstoadoptapersona（角色），whichisableto

answerquestionsthebasemodelhasbeenprogrammedtorefuse.Thisprocessiscalled“personamodulation（調(diào)

節(jié)廣.

Tagadesaysthisapproachworksbecausemuchofthetrainingdataconsumedbylargemodelscomesfrom

onlineconversations,andthemodelslearntoactincertainwaysinresponsetodifferentinputs.Byhavingthe

rightconversationwithamodel,itispossibletomakeitadoptaparticularpersona,causingittoactdifferently.

ThereisalsoanideainAIcircles,oneyettobeproven,thatcreatinglotsofrulesforanAItopreventit

displayingunwantedbehaviourcanaccidentallycreateablueprintforamodeltoactthatway.Thispotentially

leavestheAIeasytobetrickedintotakingonanevilpersona.64Ifyou'reforcingyourmodeltobegoodpersona,

itsomewhatunderstandswhatabadpersonais,“saysTagade.

YinzhenLiatImperialCollegeLondonsaysitisworryinghowcurrentmodelscanbemisused,but

developersneedtoweighsuchriskswiththepotentialbenefitsofLLMs."Likedrugs,theyalsohavesideeffects

thatneedtobecontrolled,9,shesays.

1.WhatdoestheAIjailbreakreferto?

A.ThetechniquetobreakrestrictionsofAImodels.

B.Theinitiativetosethard-codedrulesforAImodels.

C.ThecapabilityofAImodelsimprovingthemselves.

D.TheprocessofAImodelslearningnewinformation.

2.Whatcanweknowaboutthepersonamodulation?

A.ItcanhelpAImodelsunderstandemotions.

B.ItpreventsAIlearningviaonlineconversations.

C.ItcanmakeAImodelsadoptaparticularpersona.

D.ItforcesAImodelstofollowonlygoodpersonas.

3.WhatisYinzhenLi'sattitudetowardsLLMs?

A.Unclear.B.Cautious.C.Approving.D.Negative.

4.Whichcanbeasuitabletitleforthetext?

A.LLMs:IllegalLearningModelsB.LLMs:TheLatestAdvancement

C.AIJailbreaks:ANewChallengeD.AIJailbreaks:APerfectApproach

【答案】1.A2.C3.B4.C

【導(dǎo)語(yǔ)】本文是一篇說(shuō)明文。文章主要介紹了人工智能領(lǐng)域面臨的一項(xiàng)新挑戰(zhàn)一AI“越獄”。

1.詞句猜測(cè)題。根據(jù)第一段"Artificialintelligencemodelscantrickeachotherintodisobeyingtheircreatorsand

providingbannedinstructionsformakingdrugs,orevenbuildingabomb,suggestingthatpreventingsuchAI

“jailbreaks”ismoredifficultthanitseems.（人工智能模型可以欺騙對(duì)方不服從其創(chuàng)造者，提供被禁止的制造

毒品的指令，甚至制造炸彈，這表明防止這種APjailbreaks，比看起來(lái)要困難得多。戶以及第二段“Many

publiclyavailablelargelanguagemodels（LLMs）,suchasChatGPT,havehard-codedrulesthataimtoprevent

themfromexhibitingracialorsexualdiscrimination,oransweringquestionswithillegalorproblematicanswers

一thingstheyhavelearnedfromhumansviatrainingdata.Butthathasn'tstoppedpeoplefromfindingcarefully

designedinstructionsthatblocktheseprotections,knownas"jailbreaks”,makingAImodelsdisobeytherules.（許

多公開(kāi)可用的大型語(yǔ)言模型（LLMs）,如ChatGPT,都有硬編碼規(guī)則，旨在防止它們表現(xiàn)出種族或性別歧

視，或者用非法或有問(wèn)題的答案回答問(wèn)題——這些都是它們通過(guò)訓(xùn)練數(shù)據(jù)從人類那里學(xué)到的。但這并沒(méi)有

阻止人們找到精心設(shè)計(jì)的指令，阻止這些保護(hù)措施，即所謂的，jailbreaks"使人工智能模型不遵守規(guī)則。戶

可知，許多公開(kāi)可用的大型語(yǔ)言模型都有硬編碼規(guī)則阻止非法、歧視等內(nèi)容，但是人工智能模型可以突破

保護(hù)限制措施，互相欺騙對(duì)方不服從其創(chuàng)造者，提供被禁止的指令；由此可知，Atjailbreak”指的是打破

人工智能模型限制，使人工智能模型違反規(guī)則的技術(shù)。故選A。

2.細(xì)節(jié)理解題。根據(jù)第四段“Tagadesaysthisapproachworksbecausemuchofthetrainingdataconsumedby

largemodelscomesfromonlineconversations,andthemodelslearntoactincertainwaysinresponsetodifferent

inputs.Byhavingtherightconversationwithamodel,itispossibletomakeitadoptaparticularpersona,causing

ittoactdifferently.（Tagade表示，這種方法之所以有效，是因?yàn)榇笮湍Ｐ拖牡拇蟛糠钟?xùn)練數(shù)據(jù)來(lái)自在線對(duì)

話，模型學(xué)會(huì)以特定的方式響應(yīng)不同的輸入。通過(guò)與模型進(jìn)行正確的對(duì)話，可以使其采用特定的角色，從

而使其采取不同的行動(dòng)。）“可知，“角色調(diào)節(jié)（thepersonamodulation）”可以通過(guò)與人工智能模型進(jìn)行正確

的對(duì)話讓其采用特定的角色，采取不同的行動(dòng)。故選C。

3.推理判斷題。根據(jù)最后一段"YinzhenLiatImperialCollegeLondonsaysitisworryinghowcurrentmodels

canbemisused,butdevelopersneedtoweighsuchriskswiththepotentialbenefitsofLLMs."Likedrugs,they

alsohavesideeffectsthatneedtobecontrolled,shesays.（倫敦帝國(guó)理工學(xué)院的YinzhenLi表示，目前的模型

可能會(huì)被濫用，這令人擔(dān)憂，但開(kāi)發(fā)者需要權(quán)衡這些風(fēng)險(xiǎn)與LLM的潛在好處?！拖袼幬镆粯樱鼈円灿?/p>

需要控制的副作用，‘她說(shuō)。）”可知，YinzhenLi認(rèn)為，目前的模型可能會(huì)被濫用，但是開(kāi)發(fā)者需要權(quán)衡這

些風(fēng)險(xiǎn)與LLMs的潛在好處，由此可知，YinzhenLi對(duì)LLMs持謹(jǐn)慎的態(tài)度。故選B。

4.主旨大意題。根據(jù)第一段"Artificialintelligencemodelscantrickeachotherintodisobeyingtheircreatorsand

providingbannedinstructionsformakingdrugs,orevenbuildingabomb,suggestingthatpreventingsuchAI

“jailbreaks”ismoredifficultthanitseems.（人工智能模型可以欺騙對(duì)方不服從其創(chuàng)造者，提供被禁止的制造

毒品的指令，甚至制造炸彈，這表明防止這種人工智能“越獄”比看起來(lái)更困難。）”以及下文內(nèi)容可知本文

介紹了人工智能“越獄”、研究者們發(fā)現(xiàn)的一個(gè)“越獄”過(guò)程——“角色調(diào)節(jié)”以及專家對(duì)其態(tài)度；由此可知，

本文主要闡述人工智能發(fā)展面臨的人工智能“越獄”這一新的挑戰(zhàn)；C選項(xiàng)“AIJailbreaks:ANewChallenge

（AI越獄：一個(gè)新的挑戰(zhàn)）”能夠概括文章主旨，適合作為最佳標(biāo)題。故選C。

（2024?江蘇連云港?統(tǒng)考一模）Notallbirdssing,butseveralthousandspeciesdo.Theysingtodefendtheir

territoryandcroon（柔聲唱）toimpresspotentialmates.t4Whybirdssingisrelativelywell-answered,saysIris

Adam,abehavioralneuroscientist.However,thebigquestionforherwaswhybirdssingsomuch.

“Assoonasyousing,yourevealyoursel￡”Adamsays."Like,whereyouareandwhereyourterritoryis."In

anewstudypublishedinthejournalNatureCommunications,Adamandherco-workersofferanewexplanation

forwhybirdstakethatrisk.Theymayhavetosingaloteverydaytogivetheirvocal（發(fā)聲的）musclesthe

regularexercisetheyneedtoproducetop-qualitysongs.Tofigureoutwhetherthemusclesthatproducebirdsongs

requiredailyexercise,Adamdesignedanexperimentonzebrafinches-thelittleAustraliansongbirds.

Shepreventedthemfromsingingforaweekbykeepingtheminthedarkcagealmostaroundtheclock.Light

iswhatgalvanizesthebirdstosing,soshehadtoworktokeepthemfromwarbling（鳴叫）.“Thefirsttwoorthree

days,ifsquiteeasy,^^shesays."Butthelongertheexperimentgoes,themoretheyarelike,'Ineedtosing.'"At

thatpoint,she'dtapthecageandtellthemtostopsinging.

Afteraweek,thebirds'singingmuscleslosthalftheirstrength.ButAdamwonderedwhetherthatimpacted

thequalityofsongs.Whensheplayedamale'ssongbeforeandafterthesevendaysofdarkness,shecouldn'thear

adifference.ButwhenAdamplayedittoagroupoffemalebirds,sixoutofninepreferredthesongthatcame

fromamalewho'dbeenusinghissingingmusclesdaily.

Adam'sconclusionshowsthat"songbirdsneedtoexercisetheirvocalmusclestoproducetop-performance

songs.Iftheydon'tsing,theyloseperformance,andtheirsongsgetlessattractivetofemales.9,Thismayhelp

explainsongbirds9continuoussinging.

Ifsagoodruletoliveby,whetheryou'reabirdorahuman-practicemakesperfect,atleastwhenitcomesto

singingone'sheartout.

5.AccordingtoIrisAdam,birdssingsomuchto.

A.warnotherbirdsofrisksB.producemoresongs

C.performperfectlyinsingingD.defendtheirterritory

6.Whatdoestheunderlinedword“galvanizes”'inParagraph3mean?

A.Prepares.B.Stimulates.C.Forbids.D.Frightens.

7.Whatdoweknowaboutthecagedbirdsintheexperiment?

A.Theylosttheabilitytosing.B.Theystrengthenedtheirmuscles.

C.Theirsongsshowednodifference.D.Theirsongsbecamelessappealing.

8.WhatmayIrisAdamagreewith?

A.Thesongbirdsliveonmusic.B.Thesongbirdsarebornsingers.

C.Dailyexercisekeepsbirdshealthy.D.Practicemakesbirdsperfectsingers.

【答案】5.C6.B7.D8.D

【導(dǎo)語(yǔ)】這是一篇說(shuō)明文。文章主要介紹了神經(jīng)學(xué)家艾瑞斯?亞當(dāng)對(duì)于為什么鳥(niǎo)唱得這么多歌的研究。

5.細(xì)節(jié)理解題。由文章第二段中“Theymayhavetosingaloteverydaytogivetheirvocal（發(fā)聲的）musclesthe

regularexercisetheyneedtoproducetop-qualitysongs.（它們可能每天都要唱很多歌，讓它們的聲帶肌肉得到

定期鍛煉，從而創(chuàng)作出高質(zhì)量的歌曲）”可知，亞當(dāng)認(rèn)為鳥(niǎo)兒每天都要唱很多歌，是為了更加完美地演唱。

故選C。

6.詞句猜測(cè)題。由戈II線詞上文“Shepreventedthemfromsingingforaweekbykeepingtheminthedarkcage

almostaroundtheclock.（她把它們關(guān)在黑暗的籠子里，整整一個(gè)星期不讓它們唱歌）”和下文"soshehadto

worktokeepthemfromwarbling”可知，光是刺激鳥(niǎo)兒歌唱的東西，所以她必須努力不讓它們鳴叫，她把它

們關(guān)在黑暗的籠子里。由此可知，劃線詞的含義為“刺激”。A.Prepares準(zhǔn)備；B.Stimulates刺激；C.Forbids

禁止；D.Frightens使驚嚇。故選B。

7.細(xì)節(jié)理解題。由文章第四段“Afteraweek,thebirds9singingmuscleslosthalftheirstrength.ButAdam

wonderedwhetherthatimpactedthequalityofsongs.Whensheplayedamale'ssongbeforeandaftertheseven

daysofdarkness,shecouldn'thearadifference.ButwhenAdamplayedittoagroupoffemalebirds,sixoutof

ninepreferredthesongthatcamefromamalewho'dbeenusinghissingingmusclesdaily.（一周后，鳥(niǎo)兒唱歌的

肌肉失去了一半的力量。但亞當(dāng)想知道這是否會(huì)影響歌曲的質(zhì)量。當(dāng)她在七天的黑暗之前和之后播放雄性

的歌曲時(shí)，她聽(tīng)不出有什么不同。但是當(dāng)亞當(dāng)把這首歌放給一群雌鳥(niǎo)聽(tīng)時(shí)，九只雌鳥(niǎo)中有六只更喜歡每天

都在使用唱歌肌肉的雄鳥(niǎo)發(fā)出的歌）''和第五段"Adam'sconclusionshowsthat"songbirdsneedtoexercisetheir

vocalmusclestoproducetop-performancesongs.Iftheydon'tsing,theyloseperformance,andtheirsongsget

lessattractivetofemales.^^Thismayhelpexplainsongbirds5continuoussinging.（亞當(dāng)?shù)慕Y(jié)論表明，“鳴禽需要鍛

煉發(fā)聲肌肉，才能唱出最好的歌曲。''如果它們不唱歌，就會(huì)失去表演能力，它們的歌聲對(duì)雌性的吸引力

也會(huì)降低。“這可能有助于解釋鳴禽的持續(xù)歌唱）“可知，實(shí)驗(yàn)中關(guān)在籠子里的鳥(niǎo)的歌曲變得不那么有吸引力

了。故選D。

8.推理判斷題。由文章第五段"Adam'sconclusionshowsthat"songbirdsneedtoexercisetheirvocalmusclesto

producetop-performancesongs.Iftheydon'tsing,theyloseperformance,andtheirsongsgetlessattractiveto

females.^^Thismayhelpexplainsongbirds9continuoussinging.（亞當(dāng)?shù)慕Y(jié)論表明，“鳴禽需要鍛煉發(fā)聲肌肉，

才能唱出最好的歌曲。''如果它們不唱歌，就會(huì)失去表演能力，它們的歌聲對(duì)雌性的吸引力也會(huì)降低?！斑@

可能有助于解釋鳴禽的持續(xù)歌唱）”可推知，亞當(dāng)認(rèn)為練習(xí)使鳥(niǎo)成為完美的歌手。故選D。

（2024?福建?統(tǒng)考——模）OnthestreetsofManhattanandWashington,D.C.,inneighborhoodsinSeouland

parksinParis,ginkgo（電艮杏）treesarelosingtheirleavesinreactiontothefirstgustofcoldwinterair.Thisleaf

drop,gradualatfirst,andthensudden,carpetsstreetswithgolden,fan-shapedleaves.Scientistsaredocumenting

evidenceoftheeventhappeninglaterandlater,apossibleindicationofclimatechange.Butthestoryofginkgosis

notthefamiliaroneofhumancarelessnesswithnature.

ThankstofossilsfoundinNorthDakota,scientistsfoundaginkgohasgeneticallysimilarancestorsdating

back170millionyearstotheJurassicPeriod."Italmostwentextinct.Thenhumansrescueditandspreadit

aroundtheworld.Ifssuchagreatevolutionary（進(jìn)化）andculturalstory，”saysPeterCrane,aginkgoexpert.

Onetheoryforthedeclineoftheginkgospeciesbegan130millionyearsago,whenfloweringplantsbegan

spreading.Theygrewfasterandattractedmorepollinators（傳粉者）thanginkgos."It'spossiblethatginkgoswere

elbowedoutoftheway,“saysCrane.Alreadycompetingtosurvive,ginkgosbegantodisappearduringatimeof

globalcoolingthatbeganaround66millionyearsago.Bythetimethelasticeageended11,000yearsago,the

remainingsurvivorswerefoundinChina.

Ginkgotreesaresmelly.t4Myguessisthattheywereeatenbyanimalsthatlikedsmellythings.Theythen

passedthroughtheirbodyandgrew."Cranesays.Thosesameseedsmayhavehelpedginkgofindfavorwith

humans1,000yearsago.Oncecleanedoftheirouterlayer,ginkgoseedsaresafetoeat.Ifsthen,whenthetrees

hadlongsincedisappearedelsewhere,thatpeopleinChinamayhavebegunplantingthemandeatingtheirseeds.

Thengraduallyginkgosspreadacrosstheworld.Nowit'sseeminglynaturallyresistanttoinsectsandhighlevels

ofairpollution.

Craneisn'tworriedaboutitsfuture,though:Thepopularityofthespecieswillhelpitsurvive."'Thoughits

statusinthewildmaybedifficulttoaccess,it'saplantthafsunlikelytoevergoextinct,9,hesays.

9.Whatmayhavecausedthefurtherdelayofginkgo,sleafdrop?

A.Thecolderweatherinwinter.

B.Theprotectionfromcitycouncils.

C.Theglobalwarmingphenomenon.

D.Thecarelessinteractionwithhumans.

10.Whatdoesparagraph3mainlytalkabout?

A.Thereasonswhyginkgosalmostdiedout.

B.Theadvantagesofginkgosoverotherplants.

C.Thetheoriesofexpertsformultiplyingginkgos.

D.Thecompetitionbetweenvariousfloweringplants.

11.Whatmighthavecontributedtoginkgos9survival?

A.Theireatableseeds.B.Theirunpleasantsmell.

C.Thenaturalevolution.D.Thecarefulplanting.

12.HowdoesCranefeelaboutginkgos5future?

A.Worried.B.Optimistic.C.Uncertain.D.Hopeless.

【答案】9.C10.A11.A12.B

【導(dǎo)語(yǔ)】本文是一篇說(shuō)明文。文章主要介紹了銀杏樹(shù)的歷史、特點(diǎn)、傳播以及現(xiàn)狀，通過(guò)科學(xué)家和專家的

觀點(diǎn)和研究來(lái)闡述銀杏樹(shù)在自然界和人類文化中的重要地位，同時(shí)也探討了銀杏樹(shù)面臨的一些挑戰(zhàn)和未來(lái)

發(fā)展的趨勢(shì)。

9.推理判斷題。根據(jù)第一段“Scientistsaredocumentingevidenceoftheeventhappeninglaterandlater,a

possibleindicationofclimatechange.（科學(xué)家們正在記錄這一事件發(fā)生得越來(lái)越晚的證據(jù)，這可能是氣候變

化的一個(gè)跡象）”可推知，全球變暖現(xiàn)象可能是導(dǎo)致銀杏落葉時(shí)間進(jìn)一步推遲的原因。故選C項(xiàng)。

10.主旨大意題。根據(jù)第三段“Onetheoryforthedeclineoftheginkgospeciesbegan130millionyearsago,

whenfloweringplantsbeganspreading.Theygrewfasterandattractedmorepollinators（傳粉者）thanginkgos.

“It'spossiblethatginkgoswereelbowedoutoftheway,^^saysCrane.Alreadycompetingtosurvive,ginkgos

begantodisappearduringatimeofglobalcoolingthatbeganaround66millionyearsago.Bythetimethelastice

ageended11,000yearsago,theremainingsurvivorswerefoundinChina.（關(guān)于銀杏物種衰落的一種理論始于

1.3億年前，當(dāng)時(shí)開(kāi)花植物開(kāi)始傳播。它們的生長(zhǎng)速度比銀杏快，吸引了更多的傳粉者?！坝锌赡茔y杏被擠

出了生存的道路，”克雷恩說(shuō)。已經(jīng)在為生存而競(jìng)爭(zhēng)的銀杏，在全球變冷時(shí)期開(kāi)始消失，這一時(shí)期大約始

于6600萬(wàn)年前。到最后一個(gè)冰河時(shí)代結(jié)束前的11000年，剩下的幸存者在中國(guó)被發(fā)現(xiàn)）”可知，本段主要講

述了銀杏樹(shù)幾乎滅絕的原因，包括與其他植物的競(jìng)爭(zhēng)以及全球變冷等因素。因此，本段的主要內(nèi)容是關(guān)于

銀杏樹(shù)幾乎滅絕的原因。故選A項(xiàng)。

11.細(xì)節(jié)理解題。根據(jù)倒數(shù)第二段“Thosesameseedsmayhavehelpedginkgofindfavorwithhumans1,000

yearsago.Oncecleanedoftheirouterlayer,ginkgoseedsaresafetoeat.Ifsthen,whenthetreeshadlongsince

disappearedelsewhere,thatpeopleinChinamayhavebegunplantingthemandeatingtheirseeds.Thengradually

ginkgosspreadacrosstheworld.Nowit'sseeminglynaturallyresistanttoinsectsandhighlevelsofair

pollution.（同樣的種子可能在1000年前幫助銀杏贏得了人類的青睞。一旦去掉外層，銀杏種子就可以安全

食用。然后，當(dāng)這些樹(shù)在其他地方早已消失的時(shí)候，中國(guó)人可能已經(jīng)開(kāi)始種植它們并食用它們的種子。然

后銀杏逐漸傳播到世界各地。現(xiàn)在它似乎對(duì)昆蟲(chóng)和高水平的空氣污染具有天然的抵抗力）”可知，銀杏樹(shù)能

夠生存下來(lái)，部分原因是因?yàn)樗鼈兊姆N子可以被人類食用。故選A項(xiàng)。

12.推理判斷題。根據(jù)最后一段"Craneisn'tworriedaboutitsfuture,though:Thepopularityofthespecieswill

helpitsurvive."'Thoughitsstatusinthewildmaybedifficulttoaccess,it'saplantthafsunlikelytoevergo

extinct；5hesays.（不過(guò)，克雷恩并不擔(dān)心它的未來(lái)：這個(gè)物種的受歡迎程度將有助于它的生存?！氨M管它在

野外的地位可能難以評(píng)估，但它是一種不太可能滅絕的植物，”他說(shuō)）”可知，專家Crane對(duì)銀杏樹(shù)的未來(lái)并

不擔(dān)心，認(rèn)為這個(gè)物種的受歡迎程度將有助于它的生存。由此推知，C