Mova0.5 pre-release⌕

☰

home / corpora / questions / AI / How a language model learns: guessing… / t5

Test 5 · How a language model learns: guessing the next word

In a cartoon a chatbot snores in its box at night. What is wrong with that?

Mova's choice: option 2 · The right answer: option 4 · wrong ✗ · Who acts: chatbot · What happens: snore, be

This is test 5 of lesson 2 «How a language model learns: guessing the next word» of the AI course of Mova's primary school. 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

Sentence 1: In a cartoon a chatbot snores in its box at night.

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	In	in	ADP	_	3 cartoon	case
2	a	a	DET	Definite=Ind PronType=Art	3 cartoon	det
3	cartoon	cartoon	NOUN	Number=Sing	6 snores	obl
4	a	a	DET	Definite=Ind PronType=Art	5 chatbot	det
5	chatbot	chatbot	NOUN	Number=Sing	6 snores	nsubj
6	snores	snore	VERB	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	root	root
7	in	in	ADP	_	9 box	case
8	its	its	PRON	Case=Gen Gender=Neut Number=Sing Person=3 Poss=Yes PronType=Prs	9 box	nmod:poss
9	box	box	NOUN	Number=Sing	6 snores	obl
10	at	at	ADP	_	11 night	case
11	night	night	NOUN	Number=Sing	6 snores	obl
12	.	.	PUNCT	_	6 snores	punct

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	snore	present	a chatbot				night	in chatbot's box; in a cartoon	vn nonverbal_expression-40.2 «NP V PP.theme»: agent=Agent in cartoon=Theme

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.

Sentence 2: What is wrong with that?

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	What	what	PRON	PronType=Int	3 wrong	nsubj
2	is	be	AUX	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	3 wrong	cop
3	wrong	wrong	ADJ	Degree=Pos	root	root
4	with	with	ADP	_	5 that	case
5	that	that	PRON	Number=Sing PronType=Dem	3 wrong	obl
6	?	?	PUNCT	_	3 wrong	punct

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	is	present						with that; wrong property	tree rules

The options

Option 1

Nothing, bots sleep eight hours a night

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Nothing	nothing	PRON	Number=Sing PronType=Neg	4 sleep	vocative
2	,	,	PUNCT	_	1 Nothing	punct
3	bots	bot	NOUN	Number=Plur	4 sleep	nsubj
4	sleep	sleep	VERB	Mood=Ind Number=Plur Person=3 Tense=Pres VerbForm=Fin	root	root
5	eight	eight	NUM	NumForm=Word NumType=Card	6 hours	nummod
6	hours	hour	NOUN	Number=Plur	4 sleep	obl
7	a	a	DET	Definite=Ind PronType=Art	8 night	det
8	night	night	NOUN	Number=Sing	4 sleep	obl:unmarked

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	sleep	present	bots				hour; night	Nothing vocative	vn snooze-40.4-1 «NP V»: agent=Agent

Option 2 · ← Mova's choice

Chatbots only snore in the morning

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Chatbots	chatbot	NOUN	Number=Plur	3 snore	nsubj
2	only	only	ADV	_	3 snore	advmod
3	snore	snore	VERB	Mood=Ind Number=Plur Person=3 Tense=Pres VerbForm=Fin	root	root
4	in	in	ADP	_	6 morning	case
5	the	the	DET	Definite=Def PronType=Art	6 morning	det
6	morning	morning	NOUN	Number=Sing	3 snore	obl

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	snore	present	Chatbots				morning		vn nonverbal_expression-40.2 «NP V»: agent=Agent

Option 3

Bots dream but never snore

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	Bots	bot	NOUN	Number=Plur	2 dream	nsubj
2	dream	dream	VERB	Mood=Ind Number=Plur Person=3 Tense=Pres VerbForm=Fin	root	root
3	but	but	CCONJ	_	5 snore	cc
4	never	never	ADV	PronType=Neg	5 snore	advmod
5	snore	snore	VERB	VerbForm=Inf	2 dream	conj

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	dream	present	Bots						vn cognize-85 «NP V PP.topic»: agent=Agent
2	not snore	infinitive	Bots not said here						vn nonverbal_expression-40.2 «NP V»: agent=Agent

Option 4 · ✓ the right answer

A program that is not running does nothing at all

The words, one row each (CoNLL-U)

#	word	lemma	UPOS	features	head	relation
1	A	a	DET	Definite=Ind PronType=Art	2 program	det
2	program	program	NOUN	Number=Sing	7 does	nsubj
3	that	that	PRON	PronType=Rel	6 running	nsubj
4	is	be	AUX	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	6 running	aux
5	not	not	PART	Polarity=Neg	6 running	advmod
6	running	run	VERB	Tense=Pres VerbForm=Part	2 program	acl:relcl
7	does	do	VERB	Mood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin	root	root
8	nothing	nothing	PRON	Number=Sing PronType=Neg	7 does	obj
9	at	at	ADP	_	10 all	case
10	all	all	DET	PronType=Tot	8 nothing	nmod

#	verb	tense, mood	who	what, whom	to whom	where	when	other	roles by
1	not run	present · progressive	a program						vn run-51.3.2-2-1 «NP V»: agent=Theme
2	do	present	a program	nothing		all			vn act-114-1 «NP V NP»: agent=Agent nothing=Eventuality

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/t5/