Exam question 1 · How a language model learns: guessing the next word

Which of the following statements most accurately characterises the term "large language model"?

Mova's choice: option 1 · The right answer: option 1 · right ✓ · What happens: follow, characteris · To what, where: term

This is exam question 1 of lesson 2 «How a language model learns: guessing the next word» of the AI course of Mova's primary school, asked in stiff, formal language. The page gives the question's syntax tree — a dependency tree in the style of Universal Dependencies — and its pragmatics clause by clause, then the 4 options as the student saw them, each with its tree and pragmatics. Every text is parsed by Mova itself. How the student answered it, with the score of every option and why: the test at school.

The question

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Which	which	PRON	PronType=Rel	8 characterises	nsubj
2	of	of	ADP	_	5 statements	case
3	the	the	DET	Definite=Def PronType=Art	5 statements	det
4	following	follow	VERB	VerbForm=Ger	5 statements	amod
5	statements	statement	NOUN	Number=Plur	8 characterises	obl
6	most	most	ADV	Degree=Sup	7 accurately	advmod
7	accurately	accurately	ADV	_	8 characterises	advmod
8	characterises	characteris	VERB	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	root	root
9	the	the	DET	Definite=Def PronType=Art	10 term	det
10	term	term	NOUN	Number=Sing	8 characterises	obj
11	"	"	PUNCT	_	14 model	punct
12	large	large	ADJ	Degree=Pos	14 model	amod
13	language	language	NOUN	Number=Sing	14 model	compound
14	model	model	NOUN	Number=Sing	10 term	appos
15	"	"	PUNCT	_	14 model	punct
16	?	?	PUNCT	_	8 characterises	punct

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	follow	gerund							vn occur-48.3-1 «NP V»: agent=Theme
2	characteris	present		the term				of the following statements; accurately property	tree rules

One row per clause the reader finds. «not said here»: a role filled from the clause above. «roles by»: a VerbNet class and frame when one fits the verb, else the reader's rules on the tree.

The options

Option 1 · ← Mova's choice · ✓ the right answer

A very extensive program that has acquired the patterns of written language

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	A	a	DET	Definite=Ind PronType=Art	4 program	det
2	very	very	ADV	_	3 extensive	advmod
3	extensive	extensive	ADJ	Degree=Pos	4 program	amod
4	program	program	NOUN	Number=Sing	root	root
5	that	that	PRON	PronType=Rel	7 acquired	nsubj:pass
6	has	have	AUX	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	7 acquired	aux
7	acquired	acquire	VERB	Tense=Past VerbForm=Part Voice=Pass	4 program	acl:relcl
8	the	the	DET	Definite=Def PronType=Art	9 patterns	det
9	patterns	pattern	NOUN	Number=Plur	7 acquired	obj
10	of	of	ADP	_	12 language	case
11	written	write	VERB	Tense=Past VerbForm=Part Voice=Pass	12 language	amod
12	language	language	NOUN	Number=Sing	9 patterns	nmod

#	verb	tense, mood	who	what, whom	roles by
1	acquire	present · perfect · passive		a extensive program	tree rules
2	write	participle			vn scribble-25.2-1 «NP V»: agent=Agent
3	pattern		written language		nom pattern «cognition intransitive»: of=agent

Option 2

A dictionary of exceptional size printed upon paper

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	A	a	DET	Definite=Ind PronType=Art	2 dictionary	det
2	dictionary	dictionary	NOUN	Number=Sing	root	root
3	of	of	ADP	_	5 size	case
4	exceptional	exceptional	ADJ	Degree=Pos	5 size	amod
5	size	size	NOUN	Number=Sing	2 dictionary	nmod
6	printed	print	VERB	Tense=Past VerbForm=Part Voice=Pass	2 dictionary	acl
7	upon	upon	ADP	_	8 paper	case
8	paper	paper	NOUN	Number=Sing	6 printed	obl

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	print	participle		a dictionary of exceptional size not said here				upon paper	vn scribble-25.2-1 «NP V»: agent=Agent

Option 3

A physical model of the human organs of speech

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	A	a	DET	Definite=Ind PronType=Art	3 model	det
2	physical	physical	ADJ	Degree=Pos	3 model	amod
3	model	model	NOUN	Number=Sing	root	root
4	of	of	ADP	_	7 organs	case
5	the	the	DET	Definite=Def PronType=Art	7 organs	det
6	human	human	ADJ	Degree=Pos	7 organs	amod
7	organs	organ	NOUN	Number=Plur	3 model	nmod
8	of	of	ADP	_	9 speech	case
9	speech	speech	NOUN	Number=Sing	7 organs	nmod

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	model			the human organs of speech					nom model «cognition transitive»: of=patient

Option 4

A person of great stature who is fluent in many languages

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	A	a	DET	Definite=Ind PronType=Art	2 person	det
2	person	person	NOUN	Number=Sing	8 fluent	nsubj
3	of	of	ADP	_	5 stature	case
4	great	great	ADJ	Degree=Pos	5 stature	amod
5	stature	stature	NOUN	Number=Sing	2 person	nmod
6	who	who	PRON	PronType=Rel	8 fluent	nsubj
7	is	be	AUX	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	8 fluent	cop
8	fluent	fluent	ADJ	Degree=Pos	root	root
9	in	in	ADP	_	11 languages	case
10	many	many	ADJ	Degree=Pos	11 languages	amod
11	languages	language	NOUN	Number=Plur	8 fluent	obl

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	is	present	a person of great stature			many languages		fluent property	tree rules

Text: the tests of Mova's primary school, written for Mova (world/data/school/), as the course runner gave them to the student. Trees: Mova's own parse — the en model on the CPU, when the site is built — not checked by a teacher; they may be wrong. Frames: Mova's frames reader (prag), each text read on its own. Elamur Cosmic License.

This page in Markdown → https://md.mova.elamur.ai/free/en/corpora/questions/ai/2-guessing-the-next-word/e1/