Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
[論紹] Neural Text Generation in Stories Using Entity Representation as Context
Search
onizuka laboratory
July 11, 2018
Research
0
190
[論紹] Neural Text Generation in Stories Using Entity Representation as Context
弊研究室で行なったNAACL読み会の発表資料です。
onizuka laboratory
July 11, 2018
Tweet
Share
More Decks by onizuka laboratory
See All by onizuka laboratory
Phrase-Based & Neural Unsupervised Machine Translation
onilab
0
110
Tell-and-Answer: Towards Explainable Visual Question Answering using Attributes and Captions
onilab
0
65
Card-660: A Reliable Evaluation Framework for Rare Word Representation Models
onilab
0
31
A Word-Complexity Lexicon and A Neural Readability Ranking Model for Lexical Simplification
onilab
0
98
Integrating Transformer and Paraphrase Rules for Sentence Simplification
onilab
0
56
An Auto-Encoder Matching Model for Learning Utterance-Level Semantic Dependency in Dialogue Generation
onilab
0
48
Generating More Interesting Responses in Neural Conversation Models with Distributional Constraints
onilab
0
98
Modeling Multi-turn Conversation with Deep Utterance Aggregation
onilab
0
91
Learning Semantic Sentence Embeddings using Pair-wise Discriminator
onilab
0
110
Other Decks in Research
See All in Research
プロシェアリング白書2024_PROSHARING_REPORT_2024
circulation
0
660
Combating Misinformation in the age of LLMs
teacherpeterpan
0
140
Prompt Tuning から Fine Tuning への移行時期推定
icoxfog417
17
7.1k
論文紹介 DISN: Deep Implicit Surface Network for High quality Single-view 3D Reconstruction / DISN: Deep Implicit Surface Network for High quality Single-view 3D Reconstruction
nttcom
0
120
CVPR2023 EarthVision Workshopより衛星画像関連論文紹介 / Satellite Imaging Processing Papers in CVPR2023 EarthVision Workshop
nttcom
0
130
Ground Metric Learning with applications in genomics
gpeyre
0
370
Equivalence of Geodesics and Importance Weighting from the Perspective of Information Geometry
mkimura
0
140
First Authorに俺はなるっ!! IROS’23 CCC2023 FY
shota_nishiyama
0
180
ゼロからわかるリザバーコンピューティング
kurotaky
1
310
How to Perform Manual Classification for Deep Learning Using CloudCompare
kentaitakura
0
670
言語間転移学習で大規模言語モデルを賢くする
ikuyamada
7
3.5k
生成AIを用いたText to SQLの最前線
masatoto
1
2.4k
Featured
See All Featured
Writing Fast Ruby
sferik
622
60k
[RailsConf 2023] Rails as a piece of cake
palkan
28
4k
Web development in the modern age
philhawksworth
203
10k
Design by the Numbers
sachag
274
18k
What’s in a name? Adding method to the madness
productmarketing
PRO
17
2.7k
BBQ
matthewcrist
80
8.8k
The Language of Interfaces
destraynor
151
23k
The Psychology of Web Performance [Beyond Tellerrand 2023]
tammyeverts
14
1.5k
The Cost Of JavaScript in 2023
addyosmani
21
3.9k
Making Projects Easy
brettharned
109
5.5k
Principles of Awesome APIs and How to Build Them.
keavy
121
16k
Sharpening the Axe: The Primacy of Toolmaking
bcantrill
21
1.4k
Transcript
None
@ 28 0 A 2 1 ) 2 2 (
A (82 (,
C 2 A ) A2C C C GC G C
C C A A C2C 2 2AA2C E C GC C A C A C C2A C A2 GC ) A2C ,C A . ( C C A C2C 2 C GC 2 ( 2 C 2A C A2 , C 2C A C C 32 3 C 111111111111111111 2 C C C C A 2 A C 2 A
. 4 . 4 . - . - . !"
. . . . . 4 . . . #"$% . . . . 4 . . & 4 . . - . . . 4- . '( ) = max(!($% ) , #( ) , &012234" [)]) 4 . . . - . 8"
5 5 . 5 ! + 1 $ %&'(,* =
∑- .*,- /&'(,- . 5 .*,- = exp(/&'(,- 45 /&,* ) ∑ -7 exp(/&'(,-7 45 /&,* ) , /&,* . . , $ ! . , . , 5 /&'(,∗ 55 /&,* /&'(,-
. . 6 . ! . "~$(&, ()*) . ∝
exp(0123 4 5671819 :8,123 + <=8>1 4 ?(!)) . 6 6 6 66 :@ABB671 , . . 6. , . :@ABB671 6 6 6 . 6 , . 6 . C1 . 6 . 6 6
47 7 10 3 41 2211 :4 7 0 7
7 174 317 1 : 0 4 : 0 41 .1: .1 7: ! "# = % &#'( ∝ exp &#'( - ./0 , ! 2# = 3 "# = 1, &#'( ∝ exp &#'( - .56#7#895,#'( + ;<7=# - > 3 ! ?# = @ &#'(, 95A,#'( ∝ exp .B56C#D - , B &#'( ; 95A,#'( , !(G# = H|&#'(, 9JK//56#) ∝ CFSM(&#'( + .59JK//56#)
,) - - ) - ) - - -
) ,) - - ,) - ) 8 - ! " = log ' (; " = ∑+ log '(-+ ; ") " () ) ,) () ) ) -+ ) ) ) () - - - ) ) / 8 ,) ( ) - ) - ,) ) - ,) ) - ) ) ,) ) , ,) ) - ,) ( - )
) 14 5 C 5 2 B C 5 C
B C 19 2 B 45 ) 2 B C5BC ) 2 B () B57 5 CB (( B57 5 CB ( B57 5 CB C B5 C5 35B • 4 31B54 • 25 B 1 5 5 1354 0 K C 5 • 5 5 3 312 1 1 5 5 1354 0 ,K C 5 • 5 3 454 C 5 5 C9 B 3 B9BC9 7 5 C 1 4
3 3 3 : 23 1 1 E :3 3
E 1 3 ( A223 ,): E 32 2 , 3 2 , 3 E3 32 , 3 3 0 1 0A .................. 3 : 23 3 3132 3 E :3 ): E 3 2 3 3 ,): E , 3 2 , 3 , 3 , 3 ): E 3 2 3 ): E 3 2 3 3 3 3 : 23 3 A 32 33 32 1 1 A 3 2 :3 1 A 3 E 2 :3 E
4 3 2 3 43 3 • . 3 3
1 3 3 .4 . 3 3 43 . 3 4 3 43 • . 3 3 1 3 3 .4 . 14 1 43 1 4 2 11 44 • 3 .4 3 143 34 3 . 4 2 3 43 3 43
.0 1 2 2 2 2 2 2 2 4
2 01 2 1 01 2 1 0. 02 10. 5 2 2
• 3 2 2 2 2 12 • , 2
2 2 2 2 2 • 2A
2 1 2 20 21 2 24 2 2 2
42 2 2 2 2 05 0 2 2 21 2 05 2 2 02 A6 5 4 - 2 52 6 A 1 52 24 2
2 : , : 2 : , : 1 ,
A , : 1 , A 2 : 2 : : , : 2 , ,5 , , 2, ,5
• 2, , , , , , 6 6 ,1,
. 6 6 • 6 , 2, 6 , , , , , , , 6 , 6 , , 6 , , ,, • 6 , 2, 6 , , , , , 6 , 6 , 6 6 , , 6 6 6 , , , , , , , , 6 , 6 , 6 , 6 , , 6 6 2, 6 2, , ., 6 A 6 6 6 ,, 6 6 ,
• 1 1 1 1 1 , 1 1
, 7 1 1 1 , , 1 1 1 1 1 1 1 1 1 1 1 1 1 1 • , 1 1 . 1 1 , 1 , 1 7 1 1 1 1 1 , 1 7 • 1 , 7 7 7 7