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
Learning Semantic Sentence Embeddings using Pai...
Search
onizuka laboratory
October 23, 2018
Research
0
120
Learning Semantic Sentence Embeddings using Pair-wise Discriminator
弊研究室で行なったCOLING2018読み会の発表資料です。
onizuka laboratory
October 23, 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
71
Card-660: A Reliable Evaluation Framework for Rare Word Representation Models
onilab
0
33
A Word-Complexity Lexicon and A Neural Readability Ranking Model for Lexical Simplification
onilab
0
120
Integrating Transformer and Paraphrase Rules for Sentence Simplification
onilab
0
59
An Auto-Encoder Matching Model for Learning Utterance-Level Semantic Dependency in Dialogue Generation
onilab
0
55
Generating More Interesting Responses in Neural Conversation Models with Distributional Constraints
onilab
0
100
Modeling Multi-turn Conversation with Deep Utterance Aggregation
onilab
0
95
SGM: Sequence Generation Model for Multi-Label Classification
onilab
0
77
Other Decks in Research
See All in Research
EOGS: Gaussian Splatting for Efficient Satellite Image Photogrammetry
satai
4
510
Vision and LanguageからのEmbodied AIとAI for Science
yushiku
PRO
1
530
Creation and environmental applications of 15-year daily inundation and vegetation maps for Siberia by integrating satellite and meteorological datasets
satai
3
270
一人称視点映像解析の最先端(MIRU2025 チュートリアル)
takumayagi
6
3.5k
論文紹介:Not All Tokens Are What You Need for Pretraining
kosuken
0
170
SSII2025 [TS1] 光学・物理原理に基づく深層画像生成
ssii
PRO
4
4.2k
2025年度人工知能学会全国大会チュートリアル講演「深層基盤モデルの数理」
taiji_suzuki
25
18k
NLP Colloquium
junokim
1
200
2021年度-基盤研究B-研究計画調書
trycycle
PRO
0
300
20250605_新交通システム推進議連_熊本都市圏「車1割削減、渋滞半減、公共交通2倍」から考える地方都市交通政策
trafficbrain
0
770
論文読み会 SNLP2025 Learning Dynamics of LLM Finetuning. In: ICLR 2025
s_mizuki_nlp
0
200
AlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label data
satai
1
200
Featured
See All Featured
Building Flexible Design Systems
yeseniaperezcruz
328
39k
GitHub's CSS Performance
jonrohan
1032
460k
Connecting the Dots Between Site Speed, User Experience & Your Business [WebExpo 2025]
tammyeverts
8
520
Building Applications with DynamoDB
mza
96
6.6k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
229
22k
What's in a price? How to price your products and services
michaelherold
246
12k
CSS Pre-Processors: Stylus, Less & Sass
bermonpainter
358
30k
ReactJS: Keep Simple. Everything can be a component!
pedronauck
667
120k
Gamification - CAS2011
davidbonilla
81
5.4k
How To Stay Up To Date on Web Technology
chriscoyier
790
250k
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
18
1.1k
Embracing the Ebb and Flow
colly
87
4.8k
Transcript
$0-*/(:0.*,"* UI0DU -FBSOJOH4FNBOUJD4FOUFODF &NCFEEJOHT VTJOH1BJSXJTF %JTDSJNJOBUPS (3"%6"5&4$)00-PG*/'03."5*0/4$*&/$&BOE5&$)/0-0(: 04","6/*7 +6/:"5",":"."
1BQFS*/'0 #BESJ /BSBZBOB1BUSP 7JOPE,VNBS,VSNJ 4BOEFFQ,VNBS 7JOBZ/BNCPPEJSJ l-FBSOJOH4FNBOUJD4FOUFODF&NCFEEJOHT VTJOH 4FRVFOUJBM1BJSXJTF%JTDSJNJOBUPSz *O1SPDFFEJOHTPG$0-*/(
ݴ͍͑ੜϞσϧΛվྑͯͭ͠Α͍ &NCFEEJOH͕ಘΒΕΔΑ͏ʹͨ͠Α
$POUFOUT *OUSPEVDUJPO .FUIPET &YQFSJNFOUT $PODMVTJPO
#BDLHSPVOE • ҙຯ͕͍ۙจ͖ۙͮɼҙຯ͕ԕ͍จԕ͔͟ΔΑ͏ͳ จͷࢄදݱʢ&NCFEEJOHʣΛಘ͍ͨ • ݴ͍͑ੜ͕ղ͚ΔΑ͏ͳදݱΛ֫ಘ͢Εྑ͍ • 4FR4FRͰσίʔμͷηϧ͝ͱʹϩεΛܭࢉ͢ΔͨΊɼ จશମ͕ਖ਼͘͠ੜ͞ΕΔอূ͕ͳ͍ ˠࢀরจͱੜจͷҙຯతͳۙ͞Λอূ͢Δϩε͕ཉ͍͠
$POUSJCVUJPOT • ࢀরจͱੜจͷҙຯతͳۙ͞Λอূ͢Δ QBJSXJTFEJTDSJNJOBUPSMPTT HMPCBMMPTT ͷఏҊ • 4FOUJNFOU"OBMZTJTʹద༻ՄೳͰ͋Δ͜ͱΛࣔ͢ 4UBOGPSE4FOUJNFOU5SFFCBOLͰ 405"
$POUFOUT *OUSPEVDUJPO .FUIPET &YQFSJNFOUT $PODMVTJPO
.PEFM0WFSWJFX • ௨ৗͷ 4FR4FR ʹɼੜ݁Ռͱ (SPVOEUSVUIͷΤϯίʔυ݁Ռ͕ ۙ͘ͳΔΑ͏ʹֶश͢Δ 1BJSXJTF%JTDSJNJOBUPSΛՃ 1BJSXJTF%JTDSJNJOBUPS
&ODPEFS %FDPEFS -45. &ODPEFS (SPVOE5SVUI -PDBMMPTT DSPTTFOUSPQZ (MPCBMMPTT 4FOUFODF &NCFEEJOH
-PTTGVODUJPOT • -PDBMMPTT τʔΫϯ͝ͱͷ $SPTT&OUSPQZʢ௨ৗͷ 4FR4FRͷֶशͱಉ༷ʣ 1BJSXJTF%JTDSJNJOBUPS &ODPEFS
%FDPEFS -45. &ODPEFS (SPVOE5SVUI -PDBMMPTT $SPTT&OUSPQZ (MPCBMMPTT 4FOUFODF &NCFEEJOH
-PTTGVODUJPOT • (MPCBMMPTT ੜจͱࢀরจͷ &NCFEEJOHͷੵΛ࠷େԽ ੜจͱόονͷࢀরจҎ֎ͷจͷ &NCFEEJOHͷੵΛ࠷খԽ 1BJSXJTF%JTDSJNJOBUPS
&ODPEFS %FDPEFS -45. &ODPEFS (SPVOE5SVUI -PDBMMPTT DSPTTFOUSPQZ (MPCBMMPTT 4FOUFODF &NCFEEJOH
-PTTGVODUJPOT • ϩεؔʢ5PUBMʣ MPDBMMPTTͱ HMPCBMMPTTͷ 1BJSXJTF%JTDSJNJOBUPS &ODPEFS %FDPEFS
-45. &ODPEFS (SPVOE5SVUI -PDBMMPTT DSPTTFOUSPQZ (MPCBMMPTT 4FOUFODF &NCFEEJOH
$POUFOUT *OUSPEVDUJPO .FUIPET &YQFSJNFOUT $PODMVTJPO
&YQFSJNFOUTPWFSWJFX ͭͷλεΫͰ࣮ݧ • 1BSBQISBTFHFOFSBUJPO • 4FOUJNFOU"OBMZTJT
1BSBQISBTF(FOFSBUJPO • σʔληοτ 2VPSB EBUBTFU2VPSB ͷ࣭จ͔Βॏෳͨ͠ϖΞΛநग़ʢ,ʣ • ࣮ݧઃఆ ,Λ UFTUηοτɼ,Λ
WBMJEBUJPOηοτͱͯ͠༻͍Δ ੜจͱࢀরจͷྨࣅΛ #-&6ͰධՁ
1BSBQISBTF(FOFSBUJPO Ø#BTFMJOFʢ4FR4FRʣ ͱͷൺֱ %JTDSJNJOBUPSͱ &ODPEFSॏΈΛڞ༗ͨ͠ํ͕ྑ͍ σʔληοτͷαΠζʹؔΘΒͣɼ&%%-( TIBSFE ͕ͭΑ͍
1BSBQISBTF(FOFSBUJPO Øଞͷख๏ͱͷൺֱ
4FOUJNFOU"OBMZTJT • σʔληοτ 4UBOGPSE4FOUJNFOU5SFFCBOL 5SBJO, 7BMJE, 5FTU, • ࣮ݧઃఆ GJOFHSBJOFEDMBTTJGJDBUJPOUBTL
ྨ ͰධՁ \7FSZOFHBUJWF /FHBUJWF /FVUSBM 1PTJUJWF 7FSZQPTJUJWF^ ධՁࢦඪ &SSPS3BUF
4FOUJNFOU"OBMZTJT Ø݁Ռ
$POUFOUT *OUSPEVDUJPO .FUIPET &YQFSJNFOUT $PODMVTJPO
$PODMVTJPO • 4FR4FRϕʔεͷݴ͍͑ੜϞσϧʹΑΔจ &NCFEEJOHख๏ʹର͠ɼ ࢀরจͱੜจͷҙຯతͳۙ͞Λอূ͢Δ QBJSXJTFEJTDSJNJOBUPSMPTT HMPCBMMPTT ΛՃ • 1BSBQISBTF(FOFSBUJPO
4FOUJNFOU"OBMZTJTͰ 405"
͓ؾ࣋ͪʢ͋ͱͰফ͢εϥΠυʣ Ø 405"Λେʑతʹओு͢Δʹ͍Ζ͍Ζͱऑ͍ؾ͕͢Δʜ • 1BSBQISBTFଞͷσʔληοτࢼ͖͢Ͱʜ • 4FOUJNFOU 2VJDL5IPVHIUT ͱ͔ *OGFS4FOU
ͱൺֱ͖͢Ͱʜ ͳΜͰ GJOFHSBJOFE DMBTTFT ͚ͩͳͷ͔Θ͔Βͳ͍ʜ CJOBSZࡌͤͯ͘ΕͨΒ 25ͷจͱൺֱͰ͖Δͷʹʜ Ø %JTDSJNJOBUPS ػߏࣗମ 7"&ͷ ,-DPMMBQTF੍ͱ͔ʹ͑ͦ͏