Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Sign up for free
Menu
Search
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Pricing
Search
Sign in
Sign up for free
論文紹介:∞-former: Infinite Memory Transformer
Search
Sponsored
·
Your Podcast. Everywhere. Effortlessly.
Share. Educate. Inspire. Entertain. You do you. We'll handle the rest.
→
yuri
September 20, 2022
Research
440
0
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
論文紹介:∞-former: Infinite Memory Transformer
第14回最先端NLP勉強会(2022年9月26日、27日)@お茶大 発表用資料
yuri
September 20, 2022
More Decks by yuri
See All by yuri
データ指向モデリング「テキストマイニングの基礎」
yuri00
0
44
論文紹介:What In-Context Learning “Learns” In-Context: Disentangling Task Recognition and Task Learning
yuri00
0
660
論文紹介:Learning Dependency-Based Compositional Semantics
yuri00
0
170
論文紹介:What Context Features Can Transformer Language Models Use?
yuri00
0
480
Other Decks in Research
See All in Research
[最先端NLP勉強会2026] Checklists Are Better Than Reward Models For Aligning Language Models
nzw0301
1
310
CVPR2026論文紹介_VLMにとって良いvision encoderとは何か?Rethinking Model Selection in VLM Through the Lens of Gromov-Wasserstein Distance
kobayashi31
1
220
【ローカルAIに向き合う展示会vol.2】液体時間定数型モジュールを用いた オリジナルの双方向エンコーダーモデルNexteraBERT 推論速度向上検討並びにダウンストリーム評価
rikkabotan7
0
200
PGDM: Physically Guided Diffusion Model for L Downscaling
satai
3
480
Using our influence and power for patient safety
helenbevan
0
430
HAKARI-Bench - 実運用視点での情報検索モデル評価ベンチマーク
hotchpotch
1
740
ros2-perf-multihost: 分散システムにおける客観的なアーキテクチャ評価フレームワーク
takasehideki
0
260
The story of RefactoringMiner. Slow research, long-term impact
tsantalis
0
160
[BlackHatAsia2026] Hidden Telemetry: Uncovering TraceLogging ETW Providers You're Not Using (Yet)
asuna_jp
1
720
Karkada さんの論文 × 2 の紹介: (1) Closed-Form Training Dynamics Reveal Learned Features and Linear Structure in Word2Vec-like Models, (2) Symmetry in language statistics shapes the geometry of model representations
eumesy
PRO
1
740
「AIとWhyを深堀る」をAIと深堀る
iflection
0
650
CDCL を用いた MILP の厳密解法
imai448
0
260
Featured
See All Featured
A Tale of Four Properties
chriscoyier
163
24k
Visual Storytelling: How to be a Superhuman Communicator
reverentgeek
2
660
Fight the Zombie Pattern Library - RWD Summit 2016
marcelosomers
234
18k
From π to Pie charts
rasagy
1
370
So, you think you're a good person
axbom
PRO
2
2.2k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
jcasabona
35
2.8k
The Spectacular Lies of Maps
axbom
PRO
1
990
Sam Torres - BigQuery for SEOs
techseoconnect
PRO
0
540
Documentation Writing (for coders)
carmenintech
77
5.5k
The browser strikes back
jonoalderson
0
1.7k
Redefining SEO in the New Era of Traffic Generation
szymonslowik
1
420
brightonSEO & MeasureFest 2025 - Christian Goodrich - Winning strategies for Black Friday CRO & PPC
cargoodrich
3
840
Transcript
∞-former: Infinite Memory Transformer Pedro Henrique Martins, Zita Marinho, André
F. T. Martins ACL 2022 お茶大 村山友理
Prior Work • ⻑いcontext をどう扱えば良いか︖ 2 Transformer Layer 𝑋! STM
q k,v ... Transformer Layer 𝑋! STM CM q k,v ... Compressive Transformer [Rae+ 2019] Transformer-XL [Dai+ 2019]
Infinite Memory Transformer • 過去の⼊⼒系列を連続値にして扱う 3
Long-term Memory • ⼊⼒Xに畳み込み(stride=1, width=3)をし、スムージングを⾏う Lはinput size, eはembedding size •
Xを連続値 ! 𝑋(𝑡)に変換 𝑡 ∈ 0, 1 : 𝑡! = 𝑖/𝐿 𝜓 𝑡 ∈ ℝ"はN個のRBF (radial basis function) のベクトル B ∈ ℝ"×$は多変量リッジ回帰によって得られる係数⾏列 4
Long-term Memory 𝑄 = 𝑋𝑊" ∈ ℝ#×% 𝐾 = 𝐵𝑊&
∈ ℝ'×% 𝑉 = 𝐵𝑊( ∈ ℝ'×% • attention mechanism としてガウス分布を⽤いる 5
Long-term Memory • 𝑧),+ は𝑍#,-,) ∈ ℝ#×.の⾏を成す • Transformerのcontext vector
𝑍, と⾜し合わせて最終的なcontext vector 𝑍を得る 6 ← attention × value
Unbounded Memory 7 • ! 𝑋(𝑡)を圧縮 • ! 𝑋(𝑡)から𝑀個のベクトルを等間隔にサンプリング
Sticky Memories • 重要な部分のメモリを積極的に保存したほうが良いのでは︖ • 前ステップのattentionからヒストグラムを作成し、D個の等間隔なbinに分割 {𝑑/, … , 𝑑0}
• 各binについてattention probability 𝑝(𝑑1 )を計算 • 𝑝に従ってM個をサンプリング 8
Complexity • Key matrix 𝐾 は基底関数の数𝑁 だけに依存し、contextの⻑さとは無関係 • Complexityもcontextの⻑さとは独⽴ •
short-term memory も使う場合︓ • LTMのみの場合︓ • どちらもvanilla transformer より⼩さい 9
Sorting • 系列のトークンを頻度順に並べる • モデルが直近のトークンだけでなく⻑期記憶も⾒ているか調べるために、 トークンの確率分布を変化させていく • 系列が⻑くなるほど𝛼 ∈ [0,1]は0から1に徐々に増加
• vocabulary size 20 • 4,000, 8,000, 16,000トークンで実験 10
Sorting • Transformer • 3 layers • 6 attention heads
• input size L = 1,024 • memory size 2,048 • LTM (N = 1,024 basis functions) 11
Document Grounded Dialogue • CMU Document Grounded Conversation dataset (CMU-DoG)
[Zhou+ 2018] • より難しくするために、会話が始まる前にしかdocumentにアクセスできなくする • GPT-2 small + continuous LTM (N = 512 basis functions) 12
Document Grounded Dialogue 13
Document Grounded Dialogue 14
LTMのアテンションの層による違い 15
16
17
18
19
まとめ • Infinite Memory Transformer を提案 • Unbounded context •
計算量はcontextの⻑さと独⽴ • Sorting, Language modeling, Document grounded dialogue で実験 • ⻑期記憶の有⽤性を⽰した 20