ACL
#Pocket#NLP#LanguageModel#Library#KnowledgeEditing
Issue Date: 2025-05-11 EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Models, Peng Wang+, ACL24, （System Demonstrations） Commentver2.0:#1946 ... #EfficiencyImprovement#NLP
Issue Date: 2025-03-06 Full Parameter Fine-tuning for Large Language Models with Limited Resources, Lv+, ACL24, 2024.08 CommentLarge Language Models (LLMs) have revolutionized Natural Language Processing (NLP) but demand massive GPU resources for training. Lowering the thresh ... #Pocket
Issue Date: 2025-01-06 Parallel Structures in Pre-training Data Yield In-Context Learning, Yanda Chen+, arXiv24
Issue Date: 2025-05-11 EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Models, Peng Wang+, ACL24, （System Demonstrations） Commentver2.0:#1946 ... #EfficiencyImprovement#NLP
Issue Date: 2025-03-06 Full Parameter Fine-tuning for Large Language Models with Limited Resources, Lv+, ACL24, 2024.08 CommentLarge Language Models (LLMs) have revolutionized Natural Language Processing (NLP) but demand massive GPU resources for training. Lowering the thresh ... #Pocket
Issue Date: 2025-01-06 Parallel Structures in Pre-training Data Yield In-Context Learning, Yanda Chen+, arXiv24
#Survey#Pocket
Issue Date: 2025-01-06 Automated Justification Production for Claim Veracity in Fact Checking: A Survey on Architectures and Approaches, Islam Eldifrawi+, arXiv24 #Pocket
Issue Date: 2025-01-06 Legal Case Retrieval: A Survey of the State of the Art, Feng+, ACL24, 2024.08 CommentRecent years have seen increasing attention on Legal Case Retrieval (LCR), a key task in the area of Legal AI that concerns the retrieval of cases fr ... #Pocket#Dataset#Financial
Issue Date: 2025-01-06 FinTextQA: A Dataset for Long-form Financial Question Answering, Jian Chen+, ACL24 Comment@AkihikoWatanabe Do you have this dataset, please share it with me. Thank you.@thangmaster37 Thank you for your comment and I'm sorry for the late rep ... #Pocket
Issue Date: 2025-01-06 Masked Thought: Simply Masking Partial Reasoning Steps Can Improve Mathematical Reasoning Learning of Language Models, Changyu Chen+, arXiv24 Comment気になる ... #Pocket
Issue Date: 2025-01-06 A Deep Dive into the Trade-Offs of Parameter-Efficient Preference Alignment Techniques, Megh Thakkar+, arXiv24 #Pocket
Issue Date: 2025-01-06 NICE: To Optimize In-Context Examples or Not?, Pragya Srivastava+, ACL24 Comment興味深い ... #Pretraining#Pocket#InstructionTuning#PerplexityCurse
Issue Date: 2025-01-06 Instruction-tuned Language Models are Better Knowledge Learners, Zhengbao Jiang+, ACL24 Comment興味深い ... #Pocket
Issue Date: 2025-01-06 Learning to Edit: Aligning LLMs with Knowledge Editing, Yuxin Jiang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Multi-Level Feedback Generation with Large Language Models for Empowering Novice Peer Counselors, Alicja Chaszczewicz+, arXiv24 #Pocket
Issue Date: 2025-01-06 Learning Global Controller in Latent Space for Parameter-Efficient Fine-Tuning, Tan+, ACL24, 2024.08 CommentWhile large language models (LLMs) have showcased remarkable prowess in various natural language processing tasks, their training costs are exorbitan ... #Pocket
Issue Date: 2025-01-06 OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems, Chaoqun He+, arXiv24 #Pocket
Issue Date: 2025-01-06 DataDreamer: A Tool for Synthetic Data Generation and Reproducible LLM Workflows, Ajay Patel+, arXiv24 #Pocket
Issue Date: 2025-01-06 Self-Contrast: Better Reflection Through Inconsistent Solving Perspectives, Wenqi Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Llama2Vec: Unsupervised Adaptation of Large Language Models for Dense Retrieval, Li+, ACL24, 2024.08 CommentDense retrieval calls for discriminative embeddings to represent the semantic relationship between query and document. It may benefit from the using ... #Pocket#Education
Issue Date: 2025-01-06 BIPED: Pedagogically Informed Tutoring System for ESL Education, Kwon+, ACL24, 2024.08 CommentLarge Language Models (LLMs) have a great potential to serve as readily available and cost-efficient Conversational Intelligent Tutoring Systems (CIT ... #Pocket
Issue Date: 2025-01-06 Beyond Memorization: The Challenge of Random Memory Access in Language Models, Tongyao Zhu+, arXiv24 #Pocket
Issue Date: 2025-01-06 Attribute First, then Generate: Locally-attributable Grounded Text Generation, Aviv Slobodkin+, arXiv24 #Pocket
Issue Date: 2025-01-06 Can LLMs Learn from Previous Mistakes? Investigating LLMs Errors to Boost for Reasoning, Yongqi Tong+, arXiv24 #Pocket
Issue Date: 2025-01-06 Enhancing In-Context Learning via Implicit Demonstration Augmentation, Xiaoling Zhou+, arXiv24 #Pocket
Issue Date: 2025-01-06 MathGenie: Generating Synthetic Data with Question Back-translation for Enhancing Mathematical Reasoning of LLMs, Zimu Lu+, arXiv24 #Pocket
Issue Date: 2025-01-06 MELA: Multilingual Evaluation of Linguistic Acceptability, Zhang+, ACL24, 2024.08 CommentIn this work, we present the largest benchmark to date on linguistic acceptability: Multilingual Evaluation of Linguistic Acceptability—MELA, with 46 ... #Pocket
Issue Date: 2025-01-06 Time is Encoded in the Weights of Finetuned Language Models, Kai Nylund+, ACL24 #Pocket
Issue Date: 2025-01-06 Surgical Feature-Space Decomposition of LLMs: Why, When and How?, Arnav Chavan+, arXiv24 #Pocket
Issue Date: 2025-01-06 MEFT: Memory-Efficient Fine-Tuning through Sparse Adapter, Jitai Hao+, arXiv24 #Pocket
Issue Date: 2025-01-06 Benchmarking Knowledge Boundary for Large Language Models: A Different Perspective on Model Evaluation, Xunjian Yin+, arXiv24 #Pocket
Issue Date: 2025-01-06 ValueBench: Towards Comprehensively Evaluating Value Orientations and Understanding of Large Language Models, Yuanyi Ren+, arXiv24 #Pocket
Issue Date: 2025-01-06 AIR-Bench: Benchmarking Large Audio-Language Models via Generative Comprehension, Qian Yang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Self-Alignment for Factuality: Mitigating Hallucinations in LLMs via Self-Evaluation, Xiaoying Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Towards Faithful and Robust LLM Specialists for Evidence-Based Question-Answering, Tobias Schimanski+, arXiv24 #Pocket
Issue Date: 2025-01-06 AFaCTA: Assisting the Annotation of Factual Claim Detection with Reliable LLM Annotators, Jingwei+, ACL24, 2024.08 CommentWith the rise of generative AI, automated fact-checking methods to combat misinformation are becoming more and more important. However, factual claim ... #Pocket
Issue Date: 2025-01-06 Dissecting Human and LLM Preferences, Junlong Li+, arXiv24 #Pocket
Issue Date: 2025-01-06 Selene: Pioneering Automated Proof in Software Verification, Lichen Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Evaluating Intention Detection Capability of Large Language Models in Persuasive Dialogues, Sakurai+, ACL24, 2024.08 CommentWe investigate intention detection in persuasive multi-turn dialogs employing the largest available Large Language Models (LLMs).Much of the prior re ... #Pocket
Issue Date: 2025-01-06 Analyzing Temporal Complex Events with Large Language Models? A Benchmark towards Temporal, Long Context Understanding, Zhihan Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Feature-Adaptive and Data-Scalable In-Context Learning, Jiahao Li+, arXiv24 #Pocket
Issue Date: 2025-01-06 Mitigating Catastrophic Forgetting in Large Language Models with Self-Synthesized Rehearsal, Jianheng Huang+, arXiv24 #Embeddings#Pocket#Supervised-FineTuning (SFT)#RAG(RetrievalAugmentedGeneration)#LongSequence#PostTraining
Issue Date: 2025-01-06 Grounding Language Model with Chunking-Free In-Context Retrieval, Hongjin Qian+, arXiv24 CommentChunking無しでRAGを動作させられるのは非常に魅力的。一貫してかなり性能が向上しているように見える![image] ... #Pocket#LanguageModel#Evaluation#Bias
Issue Date: 2025-01-06 ConSiDERS-The-Human Evaluation Framework: Rethinking Human Evaluation for Generative Large Language Models, Aparna Elangovan+, arXiv24 #Embeddings#Pocket#Dataset#STS (SemanticTextualSimilarity)
Issue Date: 2025-01-06 Linguistically Conditioned Semantic Textual Similarity, Jingxuan Tu+, ACL24 #Pocket#NLP#LanguageModel#MoE(Mixture-of-Experts)
Issue Date: 2025-01-06 DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models, Damai+, ACL24, 2024.08 CommentIn the era of large language models, Mixture-of-Experts (MoE) is a promising architecture for managing computational costs when scaling up model param ... #Survey#Pocket#LanguageModel#MulltiModal
Issue Date: 2024-01-25 MM-LLMs: Recent Advances in MultiModal Large Language Models, Duzhen Zhang+, N_A, ACL24 Findings SummaryMM-LLMsは、コスト効果の高いトレーニング戦略を用いて拡張され、多様なMMタスクに対応する能力を持つことが示されている。本論文では、MM-LLMsのアーキテクチャ、トレーニング手法、ベンチマークのパフォーマンスなどについて調査し、その進歩に貢献することを目指している。 Comment以下、論文を斜め読みしながら、ChatGPTを通じて疑問点を解消しつつ理解した内容なので、理解が不十分な点が含まれている可能性があるので注意。 まあざっくり言うと、マルチモーダルを理解できるLLMを作りたかったら、様々なモダリティをエンコーディングして得られる表現と、既存のLLMが内部的に処理 ... #Pocket#NLP#LanguageModel#ProgressiveLearning
Issue Date: 2024-01-24 LLaMA Pro: Progressive LLaMA with Block Expansion, Chengyue Wu+, N_A, ACL24 Summary本研究では、大規模言語モデル（LLMs）の新しい事前学習後の手法を提案し、モデルの知識を効果的かつ効率的に向上させることを目指しました。具体的には、Transformerブロックの拡張を使用し、新しいコーパスのみを使用してモデルを調整しました。実験の結果、提案手法はさまざまなベンチマークで優れたパフォーマンスを発揮し、知的エージェントとして多様なタスクに対応できることが示されました。この研究は、自然言語とプログラミング言語を統合し、高度な言語エージェントの開発に貢献するものです。 Comment追加の知識を導入したいときに使えるかも?事前学習したLLaMA Blockに対して、追加のLLaMA Blockをstackし、もともとのLLaMA Blockのパラメータをfreezeした上でドメインに特化したコーパスで事後学習することで、追加の知識を挿入する。LLaMA Blockを挿入するとき ...
#NeuralNetwork#NLP#LanguageModel#Chain-of-Thought
Issue Date: 2023-04-27 Active prompting with chain-of-thought for large language models, Diao+, The Hong Kong University of Science and Technology, ACL24 Commentしっかりと読めていないが、CoT-answerが存在しないtrainingデータが存在したときに、nサンプルにCoTとAnswerを与えるだけでFew-shotの予測をtestデータに対してできるようにしたい、というのがモチベーションっぽい そのために、questionに対して、training ... #NLP#Dataset#PersonalizedGeneration
Issue Date: 2023-04-26 LaMP: When Large Language Models Meet Personalization, Selemi+, University of Massachusetts Amherst （w_ Google Research）, ACL24 Comment# 概要 Personalizationはユーザのニーズや嗜好に応えるために重要な技術で、IRやRecSysで盛んに研究されてきたが、NLPではあまり実施されてこなかった。しかし、最近のタスクで、text classificationやgeneration taskでPersonalization# ... #EfficiencyImprovement#Pocket#NLP#LanguageModel#Parallelism
Issue Date: 2025-05-16 Sequence Parallelism: Long Sequence Training from System Perspective, Li+, ACL23 Comment入力系列をチャンクに分割して、デバイスごとに担当するチャンクを決めることで原理上無限の長さの系列を扱えるようにした並列化手法。系列をデバイス間で横断する場合attention scoreをどのように計算するかが課題になるが、そのためにRing Self attentionと呼ばれるアルゴリズムを提案 ... #Pocket
Issue Date: 2025-01-06 Forgetting before Learning: Utilizing Parametric Arithmetic for Knowledge Updating in Large Language Models, Shiwen Ni+, arXiv23 #Pocket
Issue Date: 2025-01-06 Are Emergent Abilities in Large Language Models just In-Context Learning?, Sheng Lu+, arXiv23 #Pocket
Issue Date: 2025-01-06 Boosting Language Models Reasoning with Chain-of-Knowledge Prompting, Jianing Wang+, arXiv23 #Pocket
Issue Date: 2025-01-06 Exploring Memorization in Fine-tuned Language Models, Shenglai Zeng+, arXiv23 #Pocket
Issue Date: 2025-01-06 Instruction Fusion: Advancing Prompt Evolution through Hybridization, Weidong Guo+, arXiv23 #Pocket
Issue Date: 2025-01-06 Insert or Attach: Taxonomy Completion via Box Embedding, Wei Xue+, arXiv23 #Pocket
Issue Date: 2025-01-06 SPARSEFIT: Few-shot Prompting with Sparse Fine-tuning for Jointly Generating Predictions and Natural Language Explanations, Jesus Solano+, arXiv23 #Pocket
Issue Date: 2025-01-06 LoRAMoE: Alleviate World Knowledge Forgetting in Large Language Models via MoE-Style Plugin, Shihan Dou+, arXiv23 #Survey#Pocket#NLP#LanguageModel#Chain-of-Thought
Issue Date: 2025-01-06 Navigate through Enigmatic Labyrinth A Survey of Chain of Thought Reasoning: Advances, Frontiers and Future, Zheng Chu+, arXiv23 #Pocket#NLP#LanguageModel#Zero/FewShotPrompting#Chain-of-Thought
Issue Date: 2023-05-04 Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them, Mirac Suzgun+, N_A, ACL23 SummaryBIG-Bench Hard (BBH) is a suite of 23 challenging tasks that current language models have not been able to surpass human performance on. This study focuses on applying chain-of-thought prompting to BBH tasks and found that PaLM and Codex were able to surpass human performance on 10 and 17 tasks, respectively. The study also found that CoT prompting is necessary for tasks that require multi-step reasoning and that CoT and model scale interact to enable new task performance on some BBH tasks. Comment単なるfewshotではなく、CoT付きのfewshotをすると大幅にBIG-Bench-hardの性能が向上するので、CoTを使わないanswer onlyの設定はモデルの能力の過小評価につながるよ、という話らしい#InstructionTuning#read-later
Issue Date: 2023-03-30 Self-Instruct: Aligning Language Model with Self Generated Instructions, Wang+ （w_ Noah Smith）, Univesity of Washington, ACL23 CommentAlpacaなどでも利用されているself-instruction技術に関する論文# 概要 
Issue Date: 2021-09-09 Prefix-Tuning: Optimizing Continuous Prompts for Generation, Lisa+ （Percy Liang）, Stanford University, ACL21 Comment言語モデルをfine-tuningする際，エンコード時に「接頭辞」を潜在表現として与え，「接頭辞」部分のみをfine-tuningすることで（他パラメータは固定），より少量のパラメータでfine-tuningを実現する方法を提案．接頭辞を潜在表現で与えるこの方法は，GPT-3のpromptingに着 ... #NLP#ReviewGeneration
Issue Date: 2021-03-17 Unsupervised Opinion Summarization as Copycat-Review Generation, Bražinskas, ACL20 CommentOpinion summarization is the task of automatically creating summaries that reflect subjective information expressed in multiple documents, such as pr ... #Pocket#NLP#CommentGeneration#Personalization
Issue Date: 2019-09-11 Automatic Generation of Personalized Comment Based on User Profile, Zeng+, arXiv19 #NeuralNetwork#Pocket#NLP#CommentGeneration
Issue Date: 2019-08-24 Coherent Comment Generation for Chinese Articles with a Graph-to-Sequence Model, Li+ ,ACL19 #RecommenderSystems#NeuralNetwork#NaturalLanguageGeneration#Pocket#NLP#ReviewGeneration
Issue Date: 2019-08-17 Automatic Generation of Personalized Comment Based on User Profile, Zeng+, ACL19 Student Research Workshop #Pocket#NLP#DialogueGeneration
Issue Date: 2019-01-24 Training Millions of Personalized Dialogue Agents, Mazaré, ACL19 #NeuralNetwork#NLP#ReviewGeneration
Issue Date: 2019-04-12 Personalized Review Generation by Expanding Phrases and Attending on Aspect-Aware Representations, Ni+, ACL18 Comment Personalized Review Generationタスクを、uPy ... #Pocket#NLP#ReviewGeneration#Personalization
Issue Date: 2018-07-25 Personalized Review Generation by Expanding Phrases and Attending on Aspect-Aware Representations, Ni+, ACL18 #NeuralNetwork#Pocket#NLP#DialogueGeneration
Issue Date: 2018-02-08 Personalizing Dialogue Agents: I have a dog, do you have pets too?, Zhang+, ACL18 #DocumentSummarization#NeuralNetwork#Document#Supervised#Pocket#NLP
Issue Date: 2018-01-01 Coarse-to-Fine Attention Models for Document Summarization, Ling+ （with Rush）, ACL17 Workshop on New Frontiers in Summarization #Single#DocumentSummarization#NeuralNetwork#Document#Supervised#NLP#Abstractive#Admin'sPick
Issue Date: 2017-12-31 Get To The Point: Summarization with Pointer-Generator Networks, See+, ACL17 Comment解説スライド：https://www.slideshare.net/akihikowatanabe3110/get-to-the-point-summarization-with-pointergenerator-networks/1単語の生成と単語のコピーの両方を行えるハイブリッドなニューラル文書 ... #NeuralNetwork#ComputerVision#NaturalLanguageGeneration#NLP
Issue Date: 2017-12-31 Multi-Task Video Captioning with Video and Entailment Generation, Pasunuru+, ACL17 Comment解説スライド：https://www.slideshare.net/HangyoMasatsugu/hangyo-acl-paperreading2017multitask-video-captioning-with-video-and-entailment-generation/1multitas ... #NeuralNetwork#EfficiencyImprovement#NLP
Issue Date: 2017-12-31 Learning to skim text, Yu+, ACL17 Comment解説スライド：http://www.lr.pi.titech.ac.jp/~haseshun/acl2017suzukake/slides/07.pdf上記は2017年にすずかけ台で開催されたACL 2017読み会での解説スライドです。 ... #NeuralNetwork#MachineTranslation#NLP
Issue Date: 2017-12-28 Sequence-to-Dependency Neural Machine Translation, Wu+, ACL17 #PersonalizedDocumentSummarization#InteractivePersonalizedSummarization#NLP#IntegerLinearProgramming (ILP)
Issue Date: 2017-12-28 Joint Optimization of User-desired Content in Multi-document Summaries by Learning from User Feedback, P.V.S+, ACL17, 2017.08 Comment# 一言で言うと ユーザとインタラクションしながら重要なコンセプトを決め、そのコンセプトが含まれるようにILPな手法で要約を生成するPDS手法。Interactive Personalized Summarizationと似ている（似ているが引用していない、引用した方がよいのでは）。 # 手 ... #RecommenderSystems#NewsCitations#LearningToRank
Issue Date: 2018-01-01 News Citation Recommendation with Implicit and Explicit Semantics, Peng+, ACL16 Commenttarget text中に記述されているイベントや意見に対して、それらをサポートするような他のニュース記事を推薦する研究。 たとえば、target text中に「北朝鮮が先日ミサイルの発射に失敗したが...」、といった記述があったときに、このイベントについて報道しているニュース記事を推薦すると ... #Single#DocumentSummarization#NeuralNetwork#Document#Supervised#NLP#Abstractive#Admin'sPick
Issue Date: 2017-12-31 Incorporating Copying Mechanism in Sequence-to-Sequence Learning, Gu+, ACL16 Comment解説スライド：https://www.slideshare.net/akihikowatanabe3110/incorporating-copying-mechanism-in-sequene-to-sequence-learning単語のコピーと生成、両方を行えるネットワークを提案。 locati ... #Single#DocumentSummarization#NeuralNetwork#Document#Supervised#NLP#Extractive
Issue Date: 2017-12-31 Neural Summarization by Extracting Sentences and Words, Cheng+, ACL16 CommentExtractiveかつNeuralな単一文書要約ならベースラインとして使用した方がよいかも ... #NeuralNetwork#Sentence#NLP#LanguageModel
Issue Date: 2017-12-28 Larger-context language modelling with recurrent neural networks, Wang+, ACL16 Comment## 概要 通常のNeural Language Modelはsentence間に独立性の仮定を置きモデル化されているが、この独立性を排除し、preceding sentencesに依存するようにモデル化することで、言語モデルのコーパスレベルでのPerplexityが改善したという話。提案した言語 ... #NeuralNetwork#MachineTranslation#NLP#Admin'sPick
Issue Date: 2017-12-28 Pointing the unknown words, Gulcehre+, ACL16 Commentテキストを生成する際に、source textからのコピーを行える機構を導入することで未知語問題に対処した話CopyNetと同じタイミングで（というか同じconferenceで）発表 ... #NLP#LanguageModel#IJCNLP#Admin'sPick
Issue Date: 2018-03-30 Unsupervised prediction of acceptability judgements, Lau+, ACL-IJCNLP15 Comment文のacceptability（容認度）論文。 文のacceptabilityとは、native speakerがある文を読んだときに、その文を正しい文として容認できる度合いのこと。 acceptabilityスコアが低いと、Readabilityが低いと判断できる。 言語モデルをトレーニング ... #NeuralNetwork#NLP#Admin'sPick
Issue Date: 2018-02-13 Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networks, Tai+, ACL15 CommentTree-LSTM論文 ... #NeuralNetwork#Document#Embeddings#NLP
Issue Date: 2017-12-28 A hierarchical neural autoencoder for paragraphs and documents, Li+, ACL15 Comment複数文を生成(今回はautoencoder)するために、standardなseq2seq LSTM modelを、拡張したという話。 要は、paragraph/documentのrepresentationが欲しいのだが、アイデアとしては、word-levelの情報を扱うLSTM layerと ... #NaturalLanguageGeneration#Others#NLP#DataToTextGeneration
Issue Date: 2017-12-31 Comparing Multi-label Classification with Reinforcement Learning for Summarization of Time-series Data, Gkatzia+, ACL14 #Multi#DocumentSummarization#NLP#Extractive#Admin'sPick
Issue Date: 2017-12-28 Hierarchical Summarization: Scaling Up Multi-Document Summarization, Christensen+, ACL14 Comment## 概要 だいぶ前に読んだ。好きな研究。 テキストのsentenceを階層的にクラスタリングすることで、抽象度が高い情報から、関連する具体度の高いsentenceにdrill downしていけるInteractiveな要約を提案している。 ## 手法 通常のMDSでのデータセットの規模は上位に紐 ... #Multi#DocumentSummarization#NLP#Dataset#QueryBiased#Extractive#Admin'sPick
Issue Date: 2017-12-28 Query-Chain Focused Summarization, Baumel+, ACL14 Comment[Query-Chain Focused Summarization.pdf](https://github.com/AkihikoWatanabe/paper_notes/files/1590916/Query-Chain.Focused.Summarization.pdf)上記スライドは私が当時 ... #NaturalLanguageGeneration#Others#NLP#ConceptToTextGeneration#IJCNLP
Issue Date: 2017-12-31 Learning semantic correspondences with less supervision, Liang+, ACL-IJCNLP09 #MachineLearning#DomainAdaptation#NLP#Admin'sPick
Issue Date: 2017-12-31 Frustratingly easy domain adaptation, Daume, ACL07 Comment domain adaptationをする際に、Source側のFeatu ... #NaturalLanguageGeneration#RuleBased#NLP#DataToTextGeneration
Issue Date: 2017-12-31 Design of a knowledge-based report generator, Kukich, ACL83 Comment## タスク numerical stock market dataからstock market reportsを生成，我々と同様なタスク．システム名: ANA ## 手法概要 ルールベースな手法， 1) fact-generator, 2) message generator,Data2Text ... #Article#Tutorial#Slide
Issue Date: 2025-05-11 ACL 2024 参加報告, 張+, 株式会社サイバーエージェント AI Lab, 2024.08 Comment業界のトレンドを把握するのに非常に参考になる:Reasoning, KnowledgeGraph, KnowledgeEditing, DistillationPEFT, Bias, Fairness, EthicsMultimodal(QA, Benchmarking, Summ ... #Article#NeuralNetwork#NLP
Issue Date: 2021-06-10 FastSeq: Make Sequence Generation Faster, Yan+, ACL’21 CommentBART, DistilBART, T5, GPT2等のさまざまなTransformer-basedな手法で、4-9倍Inference speedを向上させる手法を提案。 ... #Article#NeuralNetwork#MachineTranslation#NLP
Issue Date: 2021-06-03 Probing Word Translations in the Transformer and Trading Decoder for Encoder Layers, ACL‘21 CommentTransformerに基づいたNMTにおいて、Encoderが入力を解釈し、Decoderが翻訳をしている、という通説を否定し、エンコーディング段階、さらにはinput embeddingの段階でそもそも翻訳が始まっていることを指摘。エンコーディングの段階ですでに翻訳が始まっているのであれば、エ ... #Article#DocumentSummarization#NeuralNetwork#NaturalLanguageGeneration#NLP
Issue Date: 2021-06-03 Incorporating Copying Mechanism in Sequence-to-Sequence Learning, Gu+, ACL’16 Comment#371 と同様コピーメカニズムを提案した論文。Joint Copy ModelやCOPYNETと呼ばれる。 次の単語が "生成" されるのか "コピー" されるのかをスコアリングし、各単語がコピーされる確率と生成される確率をMixtureした同時確率分布で表現する（ #207 等でも説明されてい解 ... #Article#DocumentSummarization#NeuralNetwork#NaturalLanguageGeneration#NLP
Issue Date: 2021-06-02 Pointing the Unknown Words, Gulcehre+, ACL’16 CommentConditional Copy Model （Pointer Softmax）を提案した論文。単語を生成する際に、語彙内の単語から生成する分布、原文の単語から生成する分布を求める。後者はattention distributionから。コピーするか否かを決める確率変数を導入し（sigmoid）、解 ... #Article#PersonalizedDocumentSummarization#DocumentSummarization#NLP#COLING
Issue Date: 2017-12-28 Automatic Text Summarization based on the Global Document Annotation, COLING-ACL, Nagao+, 1998, 1998.08 CommentPersonalized summarizationの評価はしていない。提案のみ。以下の3種類の手法を提案 keyword-based customization 関心のあるキーワードをユーザが入力し、コーパスやwordnet等の共起関係から関連語を取得し要約に利用する 文書の ...
Issue Date: 2025-01-06 Automated Justification Production for Claim Veracity in Fact Checking: A Survey on Architectures and Approaches, Islam Eldifrawi+, arXiv24 #Pocket
Issue Date: 2025-01-06 Legal Case Retrieval: A Survey of the State of the Art, Feng+, ACL24, 2024.08 CommentRecent years have seen increasing attention on Legal Case Retrieval (LCR), a key task in the area of Legal AI that concerns the retrieval of cases fr ... #Pocket#Dataset#Financial
Issue Date: 2025-01-06 FinTextQA: A Dataset for Long-form Financial Question Answering, Jian Chen+, ACL24 Comment@AkihikoWatanabe Do you have this dataset, please share it with me. Thank you.@thangmaster37 Thank you for your comment and I'm sorry for the late rep ... #Pocket
Issue Date: 2025-01-06 Masked Thought: Simply Masking Partial Reasoning Steps Can Improve Mathematical Reasoning Learning of Language Models, Changyu Chen+, arXiv24 Comment気になる ... #Pocket
Issue Date: 2025-01-06 A Deep Dive into the Trade-Offs of Parameter-Efficient Preference Alignment Techniques, Megh Thakkar+, arXiv24 #Pocket
Issue Date: 2025-01-06 NICE: To Optimize In-Context Examples or Not?, Pragya Srivastava+, ACL24 Comment興味深い ... #Pretraining#Pocket#InstructionTuning#PerplexityCurse
Issue Date: 2025-01-06 Instruction-tuned Language Models are Better Knowledge Learners, Zhengbao Jiang+, ACL24 Comment興味深い ... #Pocket
Issue Date: 2025-01-06 Learning to Edit: Aligning LLMs with Knowledge Editing, Yuxin Jiang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Multi-Level Feedback Generation with Large Language Models for Empowering Novice Peer Counselors, Alicja Chaszczewicz+, arXiv24 #Pocket
Issue Date: 2025-01-06 Learning Global Controller in Latent Space for Parameter-Efficient Fine-Tuning, Tan+, ACL24, 2024.08 CommentWhile large language models (LLMs) have showcased remarkable prowess in various natural language processing tasks, their training costs are exorbitan ... #Pocket
Issue Date: 2025-01-06 OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems, Chaoqun He+, arXiv24 #Pocket
Issue Date: 2025-01-06 DataDreamer: A Tool for Synthetic Data Generation and Reproducible LLM Workflows, Ajay Patel+, arXiv24 #Pocket
Issue Date: 2025-01-06 Self-Contrast: Better Reflection Through Inconsistent Solving Perspectives, Wenqi Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Llama2Vec: Unsupervised Adaptation of Large Language Models for Dense Retrieval, Li+, ACL24, 2024.08 CommentDense retrieval calls for discriminative embeddings to represent the semantic relationship between query and document. It may benefit from the using ... #Pocket#Education
Issue Date: 2025-01-06 BIPED: Pedagogically Informed Tutoring System for ESL Education, Kwon+, ACL24, 2024.08 CommentLarge Language Models (LLMs) have a great potential to serve as readily available and cost-efficient Conversational Intelligent Tutoring Systems (CIT ... #Pocket
Issue Date: 2025-01-06 Beyond Memorization: The Challenge of Random Memory Access in Language Models, Tongyao Zhu+, arXiv24 #Pocket
Issue Date: 2025-01-06 Attribute First, then Generate: Locally-attributable Grounded Text Generation, Aviv Slobodkin+, arXiv24 #Pocket
Issue Date: 2025-01-06 Can LLMs Learn from Previous Mistakes? Investigating LLMs Errors to Boost for Reasoning, Yongqi Tong+, arXiv24 #Pocket
Issue Date: 2025-01-06 Enhancing In-Context Learning via Implicit Demonstration Augmentation, Xiaoling Zhou+, arXiv24 #Pocket
Issue Date: 2025-01-06 MathGenie: Generating Synthetic Data with Question Back-translation for Enhancing Mathematical Reasoning of LLMs, Zimu Lu+, arXiv24 #Pocket
Issue Date: 2025-01-06 MELA: Multilingual Evaluation of Linguistic Acceptability, Zhang+, ACL24, 2024.08 CommentIn this work, we present the largest benchmark to date on linguistic acceptability: Multilingual Evaluation of Linguistic Acceptability—MELA, with 46 ... #Pocket
Issue Date: 2025-01-06 Time is Encoded in the Weights of Finetuned Language Models, Kai Nylund+, ACL24 #Pocket
Issue Date: 2025-01-06 Surgical Feature-Space Decomposition of LLMs: Why, When and How?, Arnav Chavan+, arXiv24 #Pocket
Issue Date: 2025-01-06 MEFT: Memory-Efficient Fine-Tuning through Sparse Adapter, Jitai Hao+, arXiv24 #Pocket
Issue Date: 2025-01-06 Benchmarking Knowledge Boundary for Large Language Models: A Different Perspective on Model Evaluation, Xunjian Yin+, arXiv24 #Pocket
Issue Date: 2025-01-06 ValueBench: Towards Comprehensively Evaluating Value Orientations and Understanding of Large Language Models, Yuanyi Ren+, arXiv24 #Pocket
Issue Date: 2025-01-06 AIR-Bench: Benchmarking Large Audio-Language Models via Generative Comprehension, Qian Yang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Self-Alignment for Factuality: Mitigating Hallucinations in LLMs via Self-Evaluation, Xiaoying Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Towards Faithful and Robust LLM Specialists for Evidence-Based Question-Answering, Tobias Schimanski+, arXiv24 #Pocket
Issue Date: 2025-01-06 AFaCTA: Assisting the Annotation of Factual Claim Detection with Reliable LLM Annotators, Jingwei+, ACL24, 2024.08 CommentWith the rise of generative AI, automated fact-checking methods to combat misinformation are becoming more and more important. However, factual claim ... #Pocket
Issue Date: 2025-01-06 Dissecting Human and LLM Preferences, Junlong Li+, arXiv24 #Pocket
Issue Date: 2025-01-06 Selene: Pioneering Automated Proof in Software Verification, Lichen Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Evaluating Intention Detection Capability of Large Language Models in Persuasive Dialogues, Sakurai+, ACL24, 2024.08 CommentWe investigate intention detection in persuasive multi-turn dialogs employing the largest available Large Language Models (LLMs).Much of the prior re ... #Pocket
Issue Date: 2025-01-06 Analyzing Temporal Complex Events with Large Language Models? A Benchmark towards Temporal, Long Context Understanding, Zhihan Zhang+, arXiv24 #Pocket
Issue Date: 2025-01-06 Feature-Adaptive and Data-Scalable In-Context Learning, Jiahao Li+, arXiv24 #Pocket
Issue Date: 2025-01-06 Mitigating Catastrophic Forgetting in Large Language Models with Self-Synthesized Rehearsal, Jianheng Huang+, arXiv24 #Embeddings#Pocket#Supervised-FineTuning (SFT)#RAG(RetrievalAugmentedGeneration)#LongSequence#PostTraining
Issue Date: 2025-01-06 Grounding Language Model with Chunking-Free In-Context Retrieval, Hongjin Qian+, arXiv24 CommentChunking無しでRAGを動作させられるのは非常に魅力的。一貫してかなり性能が向上しているように見える![image] ... #Pocket#LanguageModel#Evaluation#Bias
Issue Date: 2025-01-06 ConSiDERS-The-Human Evaluation Framework: Rethinking Human Evaluation for Generative Large Language Models, Aparna Elangovan+, arXiv24 #Embeddings#Pocket#Dataset#STS (SemanticTextualSimilarity)
Issue Date: 2025-01-06 Linguistically Conditioned Semantic Textual Similarity, Jingxuan Tu+, ACL24 #Pocket#NLP#LanguageModel#MoE(Mixture-of-Experts)
Issue Date: 2025-01-06 DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models, Damai+, ACL24, 2024.08 CommentIn the era of large language models, Mixture-of-Experts (MoE) is a promising architecture for managing computational costs when scaling up model param ... #Survey#Pocket#LanguageModel#MulltiModal
Issue Date: 2024-01-25 MM-LLMs: Recent Advances in MultiModal Large Language Models, Duzhen Zhang+, N_A, ACL24 Findings SummaryMM-LLMsは、コスト効果の高いトレーニング戦略を用いて拡張され、多様なMMタスクに対応する能力を持つことが示されている。本論文では、MM-LLMsのアーキテクチャ、トレーニング手法、ベンチマークのパフォーマンスなどについて調査し、その進歩に貢献することを目指している。 Comment以下、論文を斜め読みしながら、ChatGPTを通じて疑問点を解消しつつ理解した内容なので、理解が不十分な点が含まれている可能性があるので注意。 まあざっくり言うと、マルチモーダルを理解できるLLMを作りたかったら、様々なモダリティをエンコーディングして得られる表現と、既存のLLMが内部的に処理 ... #Pocket#NLP#LanguageModel#ProgressiveLearning
Issue Date: 2024-01-24 LLaMA Pro: Progressive LLaMA with Block Expansion, Chengyue Wu+, N_A, ACL24 Summary本研究では、大規模言語モデル（LLMs）の新しい事前学習後の手法を提案し、モデルの知識を効果的かつ効率的に向上させることを目指しました。具体的には、Transformerブロックの拡張を使用し、新しいコーパスのみを使用してモデルを調整しました。実験の結果、提案手法はさまざまなベンチマークで優れたパフォーマンスを発揮し、知的エージェントとして多様なタスクに対応できることが示されました。この研究は、自然言語とプログラミング言語を統合し、高度な言語エージェントの開発に貢献するものです。 Comment追加の知識を導入したいときに使えるかも?事前学習したLLaMA Blockに対して、追加のLLaMA Blockをstackし、もともとのLLaMA Blockのパラメータをfreezeした上でドメインに特化したコーパスで事後学習することで、追加の知識を挿入する。LLaMA Blockを挿入するとき ...
Issue Date: 2023-04-27 Active prompting with chain-of-thought for large language models, Diao+, The Hong Kong University of Science and Technology, ACL24 Commentしっかりと読めていないが、CoT-answerが存在しないtrainingデータが存在したときに、nサンプルにCoTとAnswerを与えるだけでFew-shotの予測をtestデータに対してできるようにしたい、というのがモチベーションっぽい そのために、questionに対して、training ... #NLP#Dataset#PersonalizedGeneration
Issue Date: 2023-04-26 LaMP: When Large Language Models Meet Personalization, Selemi+, University of Massachusetts Amherst （w_ Google Research）, ACL24 Comment# 概要 Personalizationはユーザのニーズや嗜好に応えるために重要な技術で、IRやRecSysで盛んに研究されてきたが、NLPではあまり実施されてこなかった。しかし、最近のタスクで、text classificationやgeneration taskでPersonalization# ... #EfficiencyImprovement#Pocket#NLP#LanguageModel#Parallelism
Issue Date: 2025-05-16 Sequence Parallelism: Long Sequence Training from System Perspective, Li+, ACL23 Comment入力系列をチャンクに分割して、デバイスごとに担当するチャンクを決めることで原理上無限の長さの系列を扱えるようにした並列化手法。系列をデバイス間で横断する場合attention scoreをどのように計算するかが課題になるが、そのためにRing Self attentionと呼ばれるアルゴリズムを提案 ... #Pocket
Issue Date: 2025-01-06 Forgetting before Learning: Utilizing Parametric Arithmetic for Knowledge Updating in Large Language Models, Shiwen Ni+, arXiv23 #Pocket
Issue Date: 2025-01-06 Are Emergent Abilities in Large Language Models just In-Context Learning?, Sheng Lu+, arXiv23 #Pocket
Issue Date: 2025-01-06 Boosting Language Models Reasoning with Chain-of-Knowledge Prompting, Jianing Wang+, arXiv23 #Pocket
Issue Date: 2025-01-06 Exploring Memorization in Fine-tuned Language Models, Shenglai Zeng+, arXiv23 #Pocket
Issue Date: 2025-01-06 Instruction Fusion: Advancing Prompt Evolution through Hybridization, Weidong Guo+, arXiv23 #Pocket
Issue Date: 2025-01-06 Insert or Attach: Taxonomy Completion via Box Embedding, Wei Xue+, arXiv23 #Pocket
Issue Date: 2025-01-06 SPARSEFIT: Few-shot Prompting with Sparse Fine-tuning for Jointly Generating Predictions and Natural Language Explanations, Jesus Solano+, arXiv23 #Pocket
Issue Date: 2025-01-06 LoRAMoE: Alleviate World Knowledge Forgetting in Large Language Models via MoE-Style Plugin, Shihan Dou+, arXiv23 #Survey#Pocket#NLP#LanguageModel#Chain-of-Thought
Issue Date: 2025-01-06 Navigate through Enigmatic Labyrinth A Survey of Chain of Thought Reasoning: Advances, Frontiers and Future, Zheng Chu+, arXiv23 #Pocket#NLP#LanguageModel#Zero/FewShotPrompting#Chain-of-Thought
Issue Date: 2023-05-04 Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them, Mirac Suzgun+, N_A, ACL23 SummaryBIG-Bench Hard (BBH) is a suite of 23 challenging tasks that current language models have not been able to surpass human performance on. This study focuses on applying chain-of-thought prompting to BBH tasks and found that PaLM and Codex were able to surpass human performance on 10 and 17 tasks, respectively. The study also found that CoT prompting is necessary for tasks that require multi-step reasoning and that CoT and model scale interact to enable new task performance on some BBH tasks. Comment単なるfewshotではなく、CoT付きのfewshotをすると大幅にBIG-Bench-hardの性能が向上するので、CoTを使わないanswer onlyの設定はモデルの能力の過小評価につながるよ、という話らしい#InstructionTuning#read-later
Issue Date: 2023-03-30 Self-Instruct: Aligning Language Model with Self Generated Instructions, Wang+ （w_ Noah Smith）, Univesity of Washington, ACL23 CommentAlpacaなどでも利用されているself-instruction技術に関する論文# 概要 
Issue Date: 2021-09-09 Prefix-Tuning: Optimizing Continuous Prompts for Generation, Lisa+ （Percy Liang）, Stanford University, ACL21 Comment言語モデルをfine-tuningする際，エンコード時に「接頭辞」を潜在表現として与え，「接頭辞」部分のみをfine-tuningすることで（他パラメータは固定），より少量のパラメータでfine-tuningを実現する方法を提案．接頭辞を潜在表現で与えるこの方法は，GPT-3のpromptingに着 ... #NLP#ReviewGeneration
Issue Date: 2021-03-17 Unsupervised Opinion Summarization as Copycat-Review Generation, Bražinskas, ACL20 CommentOpinion summarization is the task of automatically creating summaries that reflect subjective information expressed in multiple documents, such as pr ... #Pocket#NLP#CommentGeneration#Personalization
Issue Date: 2019-09-11 Automatic Generation of Personalized Comment Based on User Profile, Zeng+, arXiv19 #NeuralNetwork#Pocket#NLP#CommentGeneration
Issue Date: 2019-08-24 Coherent Comment Generation for Chinese Articles with a Graph-to-Sequence Model, Li+ ,ACL19 #RecommenderSystems#NeuralNetwork#NaturalLanguageGeneration#Pocket#NLP#ReviewGeneration
Issue Date: 2019-08-17 Automatic Generation of Personalized Comment Based on User Profile, Zeng+, ACL19 Student Research Workshop #Pocket#NLP#DialogueGeneration
Issue Date: 2019-01-24 Training Millions of Personalized Dialogue Agents, Mazaré, ACL19 #NeuralNetwork#NLP#ReviewGeneration
Issue Date: 2019-04-12 Personalized Review Generation by Expanding Phrases and Attending on Aspect-Aware Representations, Ni+, ACL18 Comment Personalized Review Generationタスクを、uPy ... #Pocket#NLP#ReviewGeneration#Personalization
Issue Date: 2018-07-25 Personalized Review Generation by Expanding Phrases and Attending on Aspect-Aware Representations, Ni+, ACL18 #NeuralNetwork#Pocket#NLP#DialogueGeneration
Issue Date: 2018-02-08 Personalizing Dialogue Agents: I have a dog, do you have pets too?, Zhang+, ACL18 #DocumentSummarization#NeuralNetwork#Document#Supervised#Pocket#NLP
Issue Date: 2018-01-01 Coarse-to-Fine Attention Models for Document Summarization, Ling+ （with Rush）, ACL17 Workshop on New Frontiers in Summarization #Single#DocumentSummarization#NeuralNetwork#Document#Supervised#NLP#Abstractive#Admin'sPick
Issue Date: 2017-12-31 Get To The Point: Summarization with Pointer-Generator Networks, See+, ACL17 Comment解説スライド：https://www.slideshare.net/akihikowatanabe3110/get-to-the-point-summarization-with-pointergenerator-networks/1単語の生成と単語のコピーの両方を行えるハイブリッドなニューラル文書 ... #NeuralNetwork#ComputerVision#NaturalLanguageGeneration#NLP
Issue Date: 2017-12-31 Multi-Task Video Captioning with Video and Entailment Generation, Pasunuru+, ACL17 Comment解説スライド：https://www.slideshare.net/HangyoMasatsugu/hangyo-acl-paperreading2017multitask-video-captioning-with-video-and-entailment-generation/1multitas ... #NeuralNetwork#EfficiencyImprovement#NLP
Issue Date: 2017-12-31 Learning to skim text, Yu+, ACL17 Comment解説スライド：http://www.lr.pi.titech.ac.jp/~haseshun/acl2017suzukake/slides/07.pdf上記は2017年にすずかけ台で開催されたACL 2017読み会での解説スライドです。 ... #NeuralNetwork#MachineTranslation#NLP
Issue Date: 2017-12-28 Sequence-to-Dependency Neural Machine Translation, Wu+, ACL17 #PersonalizedDocumentSummarization#InteractivePersonalizedSummarization#NLP#IntegerLinearProgramming (ILP)
Issue Date: 2017-12-28 Joint Optimization of User-desired Content in Multi-document Summaries by Learning from User Feedback, P.V.S+, ACL17, 2017.08 Comment# 一言で言うと ユーザとインタラクションしながら重要なコンセプトを決め、そのコンセプトが含まれるようにILPな手法で要約を生成するPDS手法。Interactive Personalized Summarizationと似ている（似ているが引用していない、引用した方がよいのでは）。 # 手 ... #RecommenderSystems#NewsCitations#LearningToRank
Issue Date: 2018-01-01 News Citation Recommendation with Implicit and Explicit Semantics, Peng+, ACL16 Commenttarget text中に記述されているイベントや意見に対して、それらをサポートするような他のニュース記事を推薦する研究。 たとえば、target text中に「北朝鮮が先日ミサイルの発射に失敗したが...」、といった記述があったときに、このイベントについて報道しているニュース記事を推薦すると ... #Single#DocumentSummarization#NeuralNetwork#Document#Supervised#NLP#Abstractive#Admin'sPick
Issue Date: 2017-12-31 Incorporating Copying Mechanism in Sequence-to-Sequence Learning, Gu+, ACL16 Comment解説スライド：https://www.slideshare.net/akihikowatanabe3110/incorporating-copying-mechanism-in-sequene-to-sequence-learning単語のコピーと生成、両方を行えるネットワークを提案。 locati ... #Single#DocumentSummarization#NeuralNetwork#Document#Supervised#NLP#Extractive
Issue Date: 2017-12-31 Neural Summarization by Extracting Sentences and Words, Cheng+, ACL16 CommentExtractiveかつNeuralな単一文書要約ならベースラインとして使用した方がよいかも ... #NeuralNetwork#Sentence#NLP#LanguageModel
Issue Date: 2017-12-28 Larger-context language modelling with recurrent neural networks, Wang+, ACL16 Comment## 概要 通常のNeural Language Modelはsentence間に独立性の仮定を置きモデル化されているが、この独立性を排除し、preceding sentencesに依存するようにモデル化することで、言語モデルのコーパスレベルでのPerplexityが改善したという話。提案した言語 ... #NeuralNetwork#MachineTranslation#NLP#Admin'sPick
Issue Date: 2017-12-28 Pointing the unknown words, Gulcehre+, ACL16 Commentテキストを生成する際に、source textからのコピーを行える機構を導入することで未知語問題に対処した話CopyNetと同じタイミングで（というか同じconferenceで）発表 ... #NLP#LanguageModel#IJCNLP#Admin'sPick
Issue Date: 2018-03-30 Unsupervised prediction of acceptability judgements, Lau+, ACL-IJCNLP15 Comment文のacceptability（容認度）論文。 文のacceptabilityとは、native speakerがある文を読んだときに、その文を正しい文として容認できる度合いのこと。 acceptabilityスコアが低いと、Readabilityが低いと判断できる。 言語モデルをトレーニング ... #NeuralNetwork#NLP#Admin'sPick
Issue Date: 2018-02-13 Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networks, Tai+, ACL15 CommentTree-LSTM論文 ... #NeuralNetwork#Document#Embeddings#NLP
Issue Date: 2017-12-28 A hierarchical neural autoencoder for paragraphs and documents, Li+, ACL15 Comment複数文を生成(今回はautoencoder)するために、standardなseq2seq LSTM modelを、拡張したという話。 要は、paragraph/documentのrepresentationが欲しいのだが、アイデアとしては、word-levelの情報を扱うLSTM layerと ... #NaturalLanguageGeneration#Others#NLP#DataToTextGeneration
Issue Date: 2017-12-31 Comparing Multi-label Classification with Reinforcement Learning for Summarization of Time-series Data, Gkatzia+, ACL14 #Multi#DocumentSummarization#NLP#Extractive#Admin'sPick
Issue Date: 2017-12-28 Hierarchical Summarization: Scaling Up Multi-Document Summarization, Christensen+, ACL14 Comment## 概要 だいぶ前に読んだ。好きな研究。 テキストのsentenceを階層的にクラスタリングすることで、抽象度が高い情報から、関連する具体度の高いsentenceにdrill downしていけるInteractiveな要約を提案している。 ## 手法 通常のMDSでのデータセットの規模は上位に紐 ... #Multi#DocumentSummarization#NLP#Dataset#QueryBiased#Extractive#Admin'sPick
Issue Date: 2017-12-28 Query-Chain Focused Summarization, Baumel+, ACL14 Comment[Query-Chain Focused Summarization.pdf](https://github.com/AkihikoWatanabe/paper_notes/files/1590916/Query-Chain.Focused.Summarization.pdf)上記スライドは私が当時 ... #NaturalLanguageGeneration#Others#NLP#ConceptToTextGeneration#IJCNLP
Issue Date: 2017-12-31 Learning semantic correspondences with less supervision, Liang+, ACL-IJCNLP09 #MachineLearning#DomainAdaptation#NLP#Admin'sPick
Issue Date: 2017-12-31 Frustratingly easy domain adaptation, Daume, ACL07 Comment domain adaptationをする際に、Source側のFeatu ... #NaturalLanguageGeneration#RuleBased#NLP#DataToTextGeneration
Issue Date: 2017-12-31 Design of a knowledge-based report generator, Kukich, ACL83 Comment## タスク numerical stock market dataからstock market reportsを生成，我々と同様なタスク．システム名: ANA ## 手法概要 ルールベースな手法， 1) fact-generator, 2) message generator,Data2Text ... #Article#Tutorial#Slide
Issue Date: 2025-05-11 ACL 2024 参加報告, 張+, 株式会社サイバーエージェント AI Lab, 2024.08 Comment業界のトレンドを把握するのに非常に参考になる:Reasoning, KnowledgeGraph, KnowledgeEditing, DistillationPEFT, Bias, Fairness, EthicsMultimodal(QA, Benchmarking, Summ ... #Article#NeuralNetwork#NLP
Issue Date: 2021-06-10 FastSeq: Make Sequence Generation Faster, Yan+, ACL’21 CommentBART, DistilBART, T5, GPT2等のさまざまなTransformer-basedな手法で、4-9倍Inference speedを向上させる手法を提案。 ... #Article#NeuralNetwork#MachineTranslation#NLP
Issue Date: 2021-06-03 Probing Word Translations in the Transformer and Trading Decoder for Encoder Layers, ACL‘21 CommentTransformerに基づいたNMTにおいて、Encoderが入力を解釈し、Decoderが翻訳をしている、という通説を否定し、エンコーディング段階、さらにはinput embeddingの段階でそもそも翻訳が始まっていることを指摘。エンコーディングの段階ですでに翻訳が始まっているのであれば、エ ... #Article#DocumentSummarization#NeuralNetwork#NaturalLanguageGeneration#NLP
Issue Date: 2021-06-03 Incorporating Copying Mechanism in Sequence-to-Sequence Learning, Gu+, ACL’16 Comment#371 と同様コピーメカニズムを提案した論文。Joint Copy ModelやCOPYNETと呼ばれる。 次の単語が "生成" されるのか "コピー" されるのかをスコアリングし、各単語がコピーされる確率と生成される確率をMixtureした同時確率分布で表現する（ #207 等でも説明されてい解 ... #Article#DocumentSummarization#NeuralNetwork#NaturalLanguageGeneration#NLP
Issue Date: 2021-06-02 Pointing the Unknown Words, Gulcehre+, ACL’16 CommentConditional Copy Model （Pointer Softmax）を提案した論文。単語を生成する際に、語彙内の単語から生成する分布、原文の単語から生成する分布を求める。後者はattention distributionから。コピーするか否かを決める確率変数を導入し（sigmoid）、解 ... #Article#PersonalizedDocumentSummarization#DocumentSummarization#NLP#COLING
Issue Date: 2017-12-28 Automatic Text Summarization based on the Global Document Annotation, COLING-ACL, Nagao+, 1998, 1998.08 CommentPersonalized summarizationの評価はしていない。提案のみ。以下の3種類の手法を提案 keyword-based customization 関心のあるキーワードをユーザが入力し、コーパスやwordnet等の共起関係から関連語を取得し要約に利用する 文書の ...