Nchịkọta
Least-to-Most prompting breaks a hard problem into a sequence of simpler subproblems, solving them in order so each answer feeds the next. It matters because it lets models tackle questions far harder than the examples they were shown.
Least-to-Most Prompting is part of the language-AI stack used to read, generate, classify, and transform text and speech at scale.
Ime miri emi
Least-to-Most prompting, introduced by Zhou and colleagues at Google in 2022, has two stages. First, the model is prompted to decompose a complex question into an ordered list of easier subquestions. Second, it solves those subquestions one at a time, appending each solved answer to the context so later steps can build on earlier ones. This differs from chain-of-thought, which reasons in a single pass without explicit decomposition. Nsonaazụ isiokwu ahụ siri ike n'ozuzu ya dị mfe na-esi ike: na SCAN compositional-generalization benchmark, opekempe-na-kasị mkpali edozila ọtụtụ n'ime iwu ogologo n'agbanyeghị na ihe atụ ozugbo dị mkpụmkpụ, ebe usoro echiche-nke echiche dara nke ukwuu.
Nghọta nka nka
The power comes from separating planning from execution. Decomposition produces a dependency-ordered chain so that subproblem N only relies on subproblems already solved. Azịza ọ bụla edoziziri na-ejikọta ya na ọsọ ọsọ ọsọ, na-enye ihe nlereanya ahụ nsonaazụ etiti ọ chọrọ kama ịrịọ ya ka o jide ihe niile n'otu ọtụ. Nke a na-ebelata echiche nke nzọụkwụ ọ bụla n'otu n'otu ga-arụrịrị, nke mere na ụdị n'ozuzu na ntinye aka ogologo na ike karịa otu ngosi.
Ịmụta nke kacha nta na-akpali akpali
Least-to-Most prompting breaks a hard problem into a sequence of simpler subproblems, solving them in order so each answer feeds the next. It matters because it lets models tackle questions far harder than the examples they were shown. Least-to-Most Prompting is part of the language-AI stack used to read, generate, classify, and transform text and speech at scale. Iji wulite nghọta miri emi, na-emeso obere-na-kasị ngwa ngwa dị ka ihe nlereanya na-arụ ọrụ, ọ bụghị otu njirimara: kọwaa nsonaazụ achọrọ, dokwuo anya echiche, ma kewaa ihe sistemụ nwere ike ime nke ọma na ihe ka na-achọ mkpebi ndị ọkachamara.
In practice, strong teams using Least-to-Most Prompting design prompts, retrieval, and review loops as one integrated communication system. Ha na-edepụta njirisi ịga nke ọma nke ọma, nwalee megide data ziri ezi yana usoro ọrụ, yana na-atụgharị dabere na usoro ọdịda ahụrụ karịa karịa mmeri otu oge. Nke a bụ ebe nghọta usoro ihe atụ na-atụgharị ka ọ bụrụ ike na-adịgide adịgide gafee ngwaahịa, amụma na arụmọrụ.
Usoro ọrụ asụsụ nwere ike ịga ngwa ngwa n'achụghị nkwụsi ike. N'otu oge ahụ, eziokwu ndị nwere mgbagwoju anya nwere ike tinye nwayọ nwayọ tinye akụkọ, ntinye nkwado, ma ọ bụ nsonaazụ nyocha. Ụzọ kachasị na-agbanwe agbanwe bụ ijikọ ọsọ nnwale na ịdọ aka ná ntị ọchịchị: ndị na-anya ụgbọ elu, ijide ihe akaebe, bipụta ndekọ mkpebi, na na-aga n'ihu na-emelite nchekwa dị ka omume nlereanya, atụmanya ndị ọrụ, na ihe iwu chọrọ.
Mmetụta Strategic
Usoro ọrụ asụsụ nwere ike ịga ngwa ngwa n'achụghị nkwụsi ike.
Usoro ọrụ asụsụ nwere ike ịga ngwa ngwa n'achụghị nkwụsi ike. N'ịkwanye ọkwa dị elu, a na-atụgharị nke a ka ọ bụrụ iwu arụ ọrụ enwere ike ịtụnye, oke nwe, na emume ntụlegharị ugboro ugboro ka ndị otu wee nwee ike ịbawanye ntụkwasị obi kama iwelite enweghị mgbagha.
Ọ na-agbasawanye ohere n'ofe asụsụ na ụdị nzikọrịta ozi.
Ọ na-agbasawanye ohere n'ofe asụsụ na ụdị nzikọrịta ozi. N'ịkwanye ọkwa dị elu, a na-atụgharị nke a ka ọ bụrụ iwu arụ ọrụ enwere ike ịtụnye, oke nwe, na emume ntụlegharị ugboro ugboro ka ndị otu wee nwee ike ịbawanye ntụkwasị obi kama iwelite enweghị mgbagha.
Otu dị iche iche nwere ike itinyekwu oge na ikpe ebe akpaaka na-ejikwa nkwughachi.
Otu dị iche iche nwere ike itinyekwu oge na ikpe ebe akpaaka na-ejikwa nkwughachi. N'ịkwanye ọkwa dị elu, a na-atụgharị nke a ka ọ bụrụ iwu arụ ọrụ enwere ike ịtụnye, oke nwe, na emume ntụlegharị ugboro ugboro ka ndị otu wee nwee ike ịbawanye ntụkwasị obi kama iwelite enweghị mgbagha.
Mmejuputa n'ezie n'ụwa
Solving a multi-step word problem by first listing the quantities to compute, then computing them in order
Compositional language tasks like translating long instructions into action sequences from short examples
Answering a complex research question by breaking it into sub-questions whose answers combine into the final response
Writing a program by decomposing it into helper functions solved one at a time, each reused by later steps
Usoro mmejuputa
Kachasị-na-Kacha na-akpali na omume
Solving a multi-step word problem by first listing the quantities to compute, then computing them in order.
Ịdozi nsogbu okwu ọtụtụ nzọụkwụ site na nke mbụ ịdepụta ọnụ ọgụgụ iji gbakọọ, wee gbakọọ ha n'usoro Otu egwuregwu na-enwetakarị nsonaazụ ka mma mgbe ha na-akọwapụta ọnụ ụzọ dị mma n'ihu, na-eme ka ụzọ mmadụ na-arịwanye elu maka ikpe ikpe, ma soro ma uru arụpụtaghị ihe na ụgwọ njehie na oge.
Kachasị-na-Kacha na-akpali na omume
Compositional language tasks like translating long instructions into action sequences from short examples.
Ọrụ nhazi asụsụ dị ka ịsụgharị ogologo ntuziaka ka ọ bụrụ usoro omume site na ihe atụ dị mkpirikpi Otu dị iche iche na-enwetakarị nsonaazụ ka mma mgbe ha kọwapụtara ọnụ ụzọ dị mma n'ihu, na-edobe ụzọ mmụba mmadụ maka oke ikpe, na soro ma uru nrụpụta yana ụgwọ njehie n'ime oge.
Kachasị-na-Kacha na-akpali na omume
Answering a complex research question by breaking it into sub-questions whose answers combine into the final response.
Ịza ajụjụ nyocha dị mgbagwoju anya site n'imebi ya n'ime ajụjụ ndị azịza ha jikọtara n'ime nzaghachi ikpeazụ Ndị otu na-enwetakarị nsonaazụ ka mma mgbe ha na-akọwapụta ọnụ ụzọ dị mma n'ihu, na-edebe ụzọ ịrị elu mmadụ maka ikpe ikpe, ma soro ma uru arụpụtaghị ihe na ụgwọ njehie na oge.
Kachasị-na-Kacha na-akpali na omume
Writing a program by decomposing it into helper functions solved one at a time, each reused by later steps.
Ide ihe mmemme site n'iwepụ ya n'ime ọrụ enyemaka na-edozi otu n'otu oge, nke ọ bụla ejiri usoro ọzọ mee ihe, otu na-enwetakarị nsonaazụ ka mma mgbe ha na-akọwapụta ọnụ ụzọ dị mma n'ihu, na-eme ka ụzọ mmadụ si abawanye maka ọnụ okwu ikpe, ma soro ma uru nrụpụta na ụgwọ njehie na oge.
Ihe ize ndụ & okporo ụzọ nche
Eziokwu ndị e chepụtara echepụta nwere ike jiri nwayọ tinye akụkọ, nkwado nkwado, ma ọ bụ nsonaazụ nyocha.
Mmetụta ngwa ngwa nwere ike ịmepụta nsonaazụ na-ekwekọghị ekwekọ n'ofe arịrịọ ndị yiri ya.
Enwere ike ikpughe data ederede nwere mmetụta ma ọ bụrụ na njikwa ohere adịghị ike.
Map mmejuputa
Kọwaa usoro mmepụta, ụda, na ụkpụrụ ịdịmma tupu ibugharị.
Kọwaa usoro mmepụta, ụda, na ụkpụrụ ịdịmma tupu ibugharị. Mesoo nzọụkwụ ọ bụla dị ka ọnụ ụzọ akaebe: ọ bụrụ na emezughị ụkpụrụ, kwụsịtụ mbugharị, mechie oghere ahụ, naanị wee gbasaa ojiji.
Weghachite nzaghachi site na isi mmalite ntụkwasị obi mgbe ọ bụla izi ezi dị mkpa.
Weghachite nzaghachi site na isi mmalite ntụkwasị obi mgbe ọ bụla izi ezi dị mkpa. Mesoo nzọụkwụ ọ bụla dị ka ọnụ ụzọ akaebe: ọ bụrụ na emezughị ụkpụrụ, kwụsịtụ mbugharị, mechie oghere ahụ, naanị wee gbasaa ojiji.
Debe ebe nleba anya mmadụ maka mpụta dị elu.
Debe ebe nleba anya mmadụ maka mpụta dị elu. Mesoo nzọụkwụ ọ bụla dị ka ọnụ ụzọ akaebe: ọ bụrụ na emezughị ụkpụrụ, kwụsịtụ mbugharị, mechie oghere ahụ, naanị wee gbasaa ojiji.
Sochie ụkpụrụ ọdịda ma na-azụghachi mkpali ma ọ bụ usoro ọrụ mgbe niile.
Sochie ụkpụrụ ọdịda ma na-azụghachi mkpali ma ọ bụ usoro ọrụ mgbe niile. Mesoo nzọụkwụ ọ bụla dị ka ọnụ ụzọ akaebe: ọ bụrụ na emezughị ụkpụrụ, kwụsịtụ mbugharị, mechie oghere ahụ, naanị wee gbasaa ojiji.