Textual Intelligence : Large Language Models and Their Real-World Applications

個数:
電子版価格
¥28,616
  • 電子版あり

Textual Intelligence : Large Language Models and Their Real-World Applications

  • 在庫がございません。海外の書籍取次会社を通じて出版社等からお取り寄せいたします。
    通常6~9週間ほどで発送の見込みですが、商品によってはさらに時間がかかることもございます。
    重要ご説明事項
    1. 納期遅延や、ご入手不能となる場合がございます。
    2. 複数冊ご注文の場合は、ご注文数量が揃ってからまとめて発送いたします。
    3. 美品のご指定は承りかねます。

    ●3Dセキュア導入とクレジットカードによるお支払いについて
  • 【入荷遅延について】
    世界情勢の影響により、海外からお取り寄せとなる洋書・洋古書の入荷が、表示している標準的な納期よりも遅延する場合がございます。
    おそれいりますが、あらかじめご了承くださいますようお願い申し上げます。
  • ◆画像の表紙や帯等は実物とは異なる場合があります。
  • ◆ウェブストアでの洋書販売価格は、弊社店舗等での販売価格とは異なります。
    また、洋書販売価格は、ご注文確定時点での日本円価格となります。
    ご注文確定後に、同じ洋書の販売価格が変動しても、それは反映されません。
  • 製本 Hardcover:ハードカバー版/ページ数 528 p.
  • 言語 ENG
  • 商品コード 9781394287468

Full Description

The book is a must-have resource for anyone looking to understand the complexities of generative AI, offering comprehensive insights into LLMs, effective training strategies, and practical applications.

Textual Intelligence: Large Language Models and Their Real-World Applications provides an overview of generative AI and its multifaceted applications, as well as the significance and potential of Large Language Models (LLMs), including GPT and LLaMA. It addresses the generative AI project lifecycle, challenges in existing data architectures, proposed use case planning and scope definition, model deployment, and application integration. Training LLMs, data requirements for effective LLM training, pre-training and fine-tuning processes, and navigating computational resources and infrastructure are also discussed. The volume delves into in-context learning and prompt engineering, offering strategies for crafting effective prompts, techniques for controlling model behavior and output quality, and best practices for prompt engineering.

Textual Intelligence: Large Language Models and Their Real-World Applications also discusses cost optimization strategies for LLM training, aligning models to human values, optimizing model architectures, the power of transfer learning and fine-tuning, instruction fine-tuning for precision, and parameter-efficient fine-tuning (PEFT) with adapters such as LoRA, QLoRA, and soft prompts, making it an essential guide for both beginners and industry veterans.

Readers will find this book:

Explores the real-world potential of large language models;
Introduces industry-changing AI solutions;
Provides advanced insights on AI and its models.

Audience

Industry professionals, academics, graduate students, and researchers seeking real-world solutions using generative AI.

Contents

Preface xix

Part 1: Introduction 1

1 Introduction: Overview of Generative AI and Multifaceted Applications, Significance, and Potential of LLMs 3
K. Mukheja, S. Mittal, C. Monga and S. Annam

1.1 Introduction to Generative AI and LLM 4

1.2 Applications of Generative AI 6

1.2.1 Medical 6

1.2.2 Education 7

1.2.3 Finance 7

1.3 Detail Case Study—Rise of Chatbots 9

1.3.1 Empowering Chatbots with Large Language Models 10

1.3.2 Chatbots in Medical and Healthcare Education 10

1.3.3 Chatbots in Finance 11

1.3.4 Chatbots in Tourism 11

1.4 Examples 12

1.5 Comparative Analysis of Generative AI Techniques 14

1.6 Future Scope and Potential 16

1.7 Conclusion 17

References 17

2 A Comprehensive Study of Large Language Models 21
Pawan Kumar, Anu Chaudhary, Shashank Sahu, Mradul Kumar Jain and Updesh Kumar Jaiswal

2.1 Introduction 22

2.2 Background 24

2.2.1 Tokenization 24

2.2.2 Positions Encoding 24

2.2.3 Attention in LLM 25

2.2.4 Activation Function 26

2.2.5 Data Preprocessing 26

2.2.6 Architecture Model 27

2.2.7 Pre-Training 28

2.2.8 Fine-Tuning 29

2.3 Large Language Models (LLMs) 31

2.3.1 BERT (Bidirectional Encoder Representations Transformer) 31

2.3.1.1 BERT Architecture 31

2.3.1.2 Working of BERT Model 32

2.3.1.3 Fine-Tuning in BERT 33

2.3.1.4 BERT Applications 34

2.3.1.5 Advantages of the BERT Language Model 35

2.3.1.6 Disadvantages of the BERT Language Model 35

2.3.2 ChatGPT (Chat Generative Pre-Trained Transformer) 36

2.3.2.1 ChatGPT Architecture 36

2.3.2.2 Tokenization 38

2.3.2.3 Embeddings in ChatGPT 39

2.3.2.4 Pre-Training 39

2.3.2.5 Fine-Tuning 39

2.4 Challenges and Future Directions 40

2.5 Conclusion 40

References 41

Part 2: Generative AI Project Lifecycle 45

3 A Deep Learning Methodology with Transformers LLM to Calculate the Global Temperature Difference in Recent Years 47
Ana Carolina Borges Monteiro, Reinaldo Padilha França and Rodrigo Bonacin

3.1 Introduction 48

3.2 Overview of Literature IoT 50

3.3 Overview of Literature AI 53

3.4 Methodology 56

3.5 Results 57

3.6 Discussion 61

3.7 Conclusions 63

References 64

4 Navigating the Generative AI Project Ecosystem with a Focus on Addressing Data Architecture Complexities and Strategic Model Selection for Optimal Outcomes 67
Mohammad Shabaz, Shanky Goyal, Ismail Keshta, Mukesh Soni and Vijay Kumar

4.1 Introduction 68

4.2 Literature Review 69

4.3 Proposed Method 72

4.4 Result 83

4.5 Conclusion 88

References 89

5 Generative AI Project Life Cycle—Use Case Planning and Scope Definition 93
Jyoti Rani, Pawan Kumar and Nidhi Sharma

5.1 What is Generative AI? 94

5.2 What is Artificial Intelligence? 95

5.2.1 Introduction to Generative Life Cycle 95

5.3 Generative AI on AWS 98

5.4 Why Generative AI on AWS? 99

5.5 How is Generative AI Operational? 101

5.6 Multiplicative Artificial Intelligence Interfaces 102

5.7 ChatGPT 102

5.7.1 How Does ChatGPT Work? 102

5.7.2 In What Ways is ChatGPT Being Helpful for Users? 103

5.8 What Advantages Does ChatGPT Offer? 104

5.8.1 What are ChatGPT's limitations? To What Extent is it Accurate? 105

5.9 Dall-e 106

5.9.1 How DALL-E Works 106

5.9.2 How Do You Use DALL-E? 107

5.9.3 How is DALL-E Taught? 108

5.9.4 The Prospects of ChatGPT and Generative AI 109

5.9.5 Fields that Utilize DALL-E 110

5.9.6 Advantages of Using DALL-E to Create Images 111

5.9.7 DALL-E's Effect on Image Production 112

5.9.8 Constraints with DALL-E 112

5.9.9 Examples of DALL-E's Use in the Real World 113

5.9.10 What DALL-E's Challenges Are 113

5.10 Bard 114

5.10.1 What is LaMDA? 114

5.10.2 How is Google Bard AI Used? 115

5.10.3 Google Bard AI Features 115

5.10.4 Examples and Use Cases for Google Bard AI 115

5.10.5 AI's Reach with Google Bard 116

5.10.6 Bard AI by Google vs. ChatGPT 116

5.10.7 Constraints with Google Bard AI 117

5.10.8 Important Uses of Generative AI 118

5.10.9 Creation and Manipulation of Images 118

5.11 Coding and Software 119

5.12 Making of Videos 119

5.13 Creating and Condensing Text 119

5.14 Interorganizational Cooperation 120

5.15 Enhancement of Chatbot's Performance 120

5.16 Business Exploration 121

5.17 Conclusion 121

References 122

6 Generative AI Unleashed: A Multi-Domain Journey of Successful Implementations of Large Language Models 125
Nikhil Kumar, Anurag Barthwal, Saurabh Mishra and Abhishek Jain

6.1 Introduction 126

6.1.1 Background and Motivation 126

6.1.1.1 Neural Networks and Deep Learning 127

6.1.1.2 Transformers 127

6.1.1.3 Pre-Training and Fine-Tuning 127

6.1.1.4 Scaling 127

6.1.2 Scope and Objectives 128

6.2 Literature Review 128

6.2.1 Historical Development of Generative Artificial Intelligence 129

6.2.2 Evolution of LLMs 129

6.2.3 Applications of Generative AI Across Different Domains 130

6.2.4 Challenges and Limitations in Implementing LLMs 131

6.3 Methodology 131

6.3.1 Research and Design 131

6.3.2 Methods of Data Collection 131

6.3.3 Model Selection and Training Techniques 132

6.3.4 Evaluation Measures 132

6.3.5 Ethical Considerations 132

6.4 LLM-Based Case Studies 132

6.4.1 Natural Language Generation in Healthcare 133

6.4.1.1 Case Study 1: Patient Diagnosis Support System 133

6.4.1.2 Case Study 2: Electronic Health Records Summarization 134

6.4.2 Creative Content in Media and Entertainment 134

6.4.2.1 Case Study 3: A Scriptwriting Support Tool 134

6.4.2.2 Case Study 4: Developing Virtual Characters 135

6.4.3 Language Translation and Multilingual Communication 135

6.4.3.1 Case Study 5: Multilingual Communication Platform 135

6.4.3.2 Case Study 6: Real-Time Interpretation Service 136

6.5 Results and Analysis for LLMs 136

6.5.1 Performance Evaluation of Implemented Models 136

6.5.1.1 Quantitative Metrics 137

6.5.1.2 Qualitative Analysis 139

6.5.2 Impact Assessment of LLMs Across Different Domains 139

6.5.2.1 Impact Assessment of LLMs in Healthcare 140

6.5.2.2 Impact Assessment of LLMs in Infotainment 141

6.5.2.3 Impact Assessment of LLMs in Language Translation 142

6.5.3 User Feedback and Acceptance 143

6.5.3.1 A/B Testing: Choice as a Coping Strategy 144

6.5.3.2 Surveys: Capturing Broad Feedback 144

6.5.3.3 User Interviews: Getting Into the Weeds of UX 144

6.5.4 Comparison with Existing Systems 145

6.6 Discussion 145

6.6.1 Understanding the Successful Implementation of LLMs 145

6.6.1.1 Multimodal Generative AI: Unleashing the Power of Many Data Types 146

6.6.2 Challenges and Limitations 147

6.6.3 Ethical Implications and Responsible AI Practices 148

6.6.4 Future Directions and Emerging Trends 149

6.6.4.1 LLMs: A Powerful Tool, But One That Demands Careful Consideration for Society 150

6.7 Conclusion 151

References 152

Appendix 155

Glossary 155

7 Misbehaving AI Models and AI Interaction Issues with Humans 157
Nishi Gupta and Shikha Gupta

7.1 Introduction 158

7.2 Literature Review 160

7.3 Misbehaving AI Models 162

7.3.1 Causes of Misbehaving AI Models 162

7.3.2 Consequences of Misbehaving AI Models 164

7.3.3 Mitigation Strategies That Can Be Employed to Address Misbehaving AI Models 167

7.4 Human Interaction with AI models 168

7.4.1 Human Interaction Issues with AI Models 168

7.4.2 Laws Made to Deal with Misbehaving AI Models 169

7.4.3 The Importance of Ongoing Research and Development in Addressing Misbehaving AI Models 171

7.5 Conclusion 173

References 174

8 Decoding Potential of ChatGPT: A Comprehensive Exploration of AI Generated Contents and Challenges 177
Anju Kaushik and Anil Kaushik

8.1 Introduction 178

8.2 Chapter Organization 179

8.3 ChatGPT Popularity Statistics 179

8.4 Implementation and Work Flow of ChatGPT 180

8.5 ChatGPT Key Characteristics in Present Scenario 182

8.6 Potential Challenges 186

8.7 Security Threats in ChatGPT 187

8.8 ChatGPT's Privacy Risks 189

8.9 Ethical Concern 192

8.10 Computer Ethics Challenges Raised by ChatGPT 194

8.11 Limitation of ChatGPT 195

8.12 Balance Between Human Knowledge and AI-Supported Innovation 196

8.13 Future Challenges 197

8.14 Conclusion 197

References 198

9 Economizing Large Language Model Training and Alignment with Human Values through Cost Effective Architectures and Transfer Learning Techniques 201
Mohammed Wasim Bhatt, Rubal Jeet, Mukesh Soni, Haewon Byeon and Vishal Sagar

9.1 Introduction 202

9.2 Literature Survey 203

9.3 Proposed Method 205

9.4 Results 216

9.5 Discussion 219

9.6 Conclusion 219

References 220

Part 3: In-Context Learning/Prompt Engineering 223

10 From Prompts to Performance: Innovations in Context Learning 225
Amandeep Sharma, Prince Kumar and Shashank Dhamija

10.1 The Art of Prompt Engineering: A Deep Dive 226

10.1.1 Core Definitions and Key Concepts of Prompt Engineering 226

10.1.1.1 Significance of Prompt Engineering 226

10.1.1.2 Fundamental Components of a Prompt 226

10.1.1.3 Prompt Engineering's Technical Aspects 228

10.2 Strategies for Crafting Effective Prompts 229

10.3 Techniques for Controlling the Model Behavior and Output 245

10.4 Best Practices for Prompt Engineering 246

10.4.1 Prompt Engineering Principles 247

10.4.2 Structured Procedure Behind Prompt Engineering 247

10.4.3 Prompt Engineering Use Cases and Applications 248

References 250

Part 4: LangChain Framework 253

11 Introduction to LangChain Framework 255
Deepti Goyal and Amita Gautam

11.1 Introduction of LangChain Framework 256

11.2 Large Language Model (LLM) 258

11.3 What Do You Mean by Chains in LangChain Framework 260

11.3.1 Various Types of Chains 260

11.3.1.1 LLMChain 261

11.3.1.2 Router Chain 261

11.3.1.3 Sequential Chain 262

11.4 Why LangChain Framework is Important 263

11.5 Main Components of LangChain Framework 264

11.5.1 Large Language Model (LLM) 264

11.5.2 Prompt Template 265

11.5.2.1 Indexes 265

11.5.2.2 Retriever 265

11.5.2.3 Parsers for Output 265

11.5.2.4 Vector Store 266

11.5.2.5 Agents 266

11.5.2.6 Memory 266

11.5.2.7 Chain 267

11.6 Feature of LangChain Framework 267

11.6.1 Scalability 267

11.6.2 Improved Usability 267

11.6.3 Adaptability 267

11.6.4 Extension 267

11.6.5 External Integrations 268

11.6.6 Thriving Community 268

11.6.7 Flexibility Across Zones 268

11.6.8 Integrations 268

11.6.9 Standardized Interfaces 268

11.6.10 Prompt Management and Optimization 268

11.6.11 Visualization and Experimentation 268

11.7 How to Install 269

11.7.1 Steps to Develop an Application in LangChain Framework 270

11.7.1.1 Describe the Use Case 270

11.7.1.2 Develop Functionality 270

11.7.1.3 Tailor the Functionality 270

11.7.1.4 Optimizing LLMs 270

11.7.1.5 Data Purification 270

11.7.1.6 Experimenting 271

11.7.2 Build a New Application with LangChain Framework 271

11.8 Real World Applications with LangChain Framework 272

11.8.1 LangSmith 272

11.8.2 Chatbots 272

11.8.3 Automated Blog Outlines 272

11.8.4 Integration with MongoDB Atlas 272

11.8.5 Medical Care 272

11.8.6 Help with Coding 273

11.8.7 Creating Condensed Content 273

11.9 Integration of LangChain Framework 273

11.10 Creating a Prompt in LangChain Framework 274

11.10.1 Types of LangChain Prompts 275

11.10.2 Prompt Template 275

11.10.3 Few_Shot_Prompt_Template 276

11.10.4 Chat_Prompt_Template 276

11.11 Future of LangChain Framework with AI Enabled Tools 278

11.11.1 ChatGPT and Chatbots 278

11.11.2 AI-Powered Text Categorization Tools 278

11.11.3 False References 279

11.12 Limitation of LangChain Framework 279

11.13 Alternative Technologies Apart from LangChain Framework Used in 2024 280

11.13.1 Auto-GPT: Bringing AI Agent Development to New Heights 280

11.13.2 Prompt_Chainer 281

11.13.3 Auto_Chain 282

11.13.4 AgentGPT: Unleashing the Power of Autonomous AI Agents 282

11.13.5 BabyAGI: A Glimpse Into the Future of Task-Driven AI 283

11.13.6 SimpleaiChat 283

11.13.7 GradientJ: Building LLM-Powered Applications with Ease 284

11.14 Conclusion 284

References 285

12 LangChain: Simplifying Development with Language Models 287
Sangeetha Annam, Merry Saxena, Ujjwal Kaushik and Shikha Mittal

12.1 Introduction 288

12.2 Phases and Characteristics of LLM Application 289

12.3 Components and Key Elements of LLM 290

12.4 Types and Architecture of LLM 293

12.5 Benefits and Approaches of LLM 296

12.6 Building an LLM Application 299

12.7 Use Cases 300

References 302

13 Addressing Ethical Challenges in LLMs: Bias and Misinformation 305
Pummy Dhiman and Amandeep Kaur

13.1 Introduction 305

13.2 LLM Evolution Tree 308

13.2.1 Bert 309

13.2.2 Gpt 311

13.3 Types of LLMs 313

13.4 Limitations of LLMs 314

13.5 Factors Contributing to Bias and Misinformation Generation 316

13.6 Methods to Address Bias and Misinformation 317

13.7 Conclusion 319

References 320

Part 5: LLM-Powered Applications 323

14 LegalEase: Application Development with LangChain Framework 325
Nidhi Malik, Lakshita Chhikara, Abhilakshay and Ambika Thakur

14.1 Introduction 325

14.1.1 Large Language Model 326

14.1.2 General Architecture 327

14.1.3 Examples of LLMs 329

14.1.4 Benefits 329

14.1.5 Industry Applications 330

14.2 LangChain 331

14.2.1 Key Features of LangChain 331

14.2.2 Key Components 333

14.2.3 Who Should Explore 335

14.3 Example of Application Development 335

14.3.1 Key Features 336

14.3.2 Purpose and Benefits 336

14.4 Development Steps 337

14.4.1 Libraries and Imports 337

14.4.2 Environment Setup 340

14.4.3 Data Collection 341

14.4.4 User Interface Setup 342

14.4.5 Document Summarization 343

14.4.6 Querying the Document 355

14.5 Conclusion 362

References 363

15 Unveiling the Potential of Massive Language Models in Software Engineering: Exploring Opportunities, Addressing Risks, and Comprehending Implications 365
Mitali Chugh

15.1 Introduction 366

15.2 Harnessing the Power: Abilities of Large Language Models 367

15.3 Navigating Challenges: Risks and Ethical Considerations 369

15.4 Ethical Application: Strategies and Frameworks 371

15.5 Establishing Ethical Frameworks for Accountability 372

15.6 Collaborative Standards: Industry and Research Collaboration 373

15.7 Transformative Effects: Broader Implications in Software Engineering 375

15.8 Shaping the Future: Prospective Directions of Large Language Models 377

15.9 Conclusion 378

References 379

16 Multidimensional Impacts of Generative AI and an In-Depth Analysis of LLMs with Their Expanding Horizons in Technology and Society 383
Rubal Jeet, Mohammed Wasim Bhatt, Maher Ali Rusho, Aadam Quraishi and Mahesh Manchanda

16.1 Introduction 384

16.2 Literature Review 386

16.3 Proposed Methodology 389

16.4 Results 402

16.5 Conclusion 408

References 409

Part 6: Responsible AI 413

17 Responsible AI: Ethical Considerations in Generative AI 415
Kamal Kumar and Poonam

17.1 Introduction 416

17.1.1 Defining Generative AI 416

17.1.2 Distinguishing Machine Learning Approaches 417

17.1.3 Brief History and Recent Breakthroughs 417

17.1.4 Overview of Key Generative Architectures and Techniques 420

17.1.4.1 Autoregressive Models 420

17.1.4.2 Generative Adversarial Networks (GANs) 420

17.1.4.3 VariationalAutoencoders (VAEs) 420

17.1.4.4 Diffusion Models 421

17.1.4.5 Self-Supervised, Meta and Multi-Task Learning 422

17.1.5 Promising Applications and Benefits 422

17.2 Key Ethical Considerations, Risks, and Challenges 423

17.2.1 Societal Biases and Unfair Representational Harms 423

17.2.2 Truth Manipulation and Attribution Difficulties 424

17.2.3 Violations of Consent, Privacy, and Agency 424

17.2.4 Misuse Potentials Across Fraud, Deceit, and Sabotage 424

17.2.5 Broader Societal Impacts on Economics, Culture and Psychology 425

17.3 Guiding Principles and Frameworks for Responsible Generative AI 425

17.3.1 Transparency 426

17.3.2 Justice, Fairness, and Inclusion 426

17.3.3 Non-Maleficence 426

17.3.4 Responsibility and Accountability 426

17.3.5 Privacy and Data Protection 426

17.4 Governance Strategies for Trustworthy Generative AI Innovation 427

17.4.1 AI Ethics Guidelines and Organizational Policies 427

17.4.2 Laws, Regulations, and Dynamic Governance Complexities 427

17.4.3 Technical Approaches to Fairness, Transparency and Control 427

17.4.4 Stakeholder Participation and Public Discourse Ethics 428

17.5 Recommendations for Key Generative AI Stakeholders 428

17.5.1 Guidelines for Technology Researchers and Developers 428

17.5.2 Strategies for Organizations, Platforms, and Corporations 429

17.5.3 Ethical Governance Strategies for Organizations 429

17.5.4 Policy Options for Governments and Lawmakers 429

17.5.5 Priorities for Broader Industry Governance Entities 430

17.5.6 Considerations for Civil Society Groups, Activists, and General Public 430

17.5.7 The Impact of Generative AI Like ChatGPT on Education 430

Significant Risks and Difficulties to Surmount 431

Research Priorities for the Future 431

17.6 Conclusions 432

References 433

18 From Prototyping to Deployment: Human-Centered Design Practices in Responsible AI Innovation 435
Jyoti Snehi, Manish Snehi, Isha Kansal and Vikas Khullar

18.1 Introduction 436

18.2 Literature Review 441

Overview of Human-Centered Design Principles 443

Responsible AI 447

Gaps in Existing Research 451

Methodology 452

Research Design 452

Rationale for Qualitative Approach 452

Human-Centered Design in AI Prototyping 456

Distinctions and Issues 456

User Research and Personas 456

Early-Phase Prototyping 457

Iterative Design and Feedback Loops 457

Ethical Considerations in AI Prototyping 458

Identifying Ethical Challenges 458

Incorporating Ethical Guidelines Into Prototyping 458

Case Studies of Ethical AI Prototyping 459

From Prototyping to Development 459

Transitioning From Prototype to Full Development 460

Ensuring Consistency in HCD Practices 460

Collaboration Across Multidisciplinary Teams 461

Tools and Techniques for Managing Development Phases 461

Human-Centered Design in AI Deployment 462

Challenges and Solutions 463

Common Challenges in Implementing HCD in AI 463

Solutions and Best Practices 465

Lessons Learned From Case Studies 467

Framework for Human-Centered and Responsible AI 469

18.3 Conclusion 471

References 472

19 Toward Accurate Abbreviation Disambiguation in Medical Texts: A Comparative Study of AI Models 475
A. Pandey and M. Saini

19.1 Introduction 476

19.2 Related Work 477

19.3 Datasets 479

19.4 Methodology 480

19.4.1 Data Collection 481

19.4.2 Pre-Processing 481

19.4.3 Vector Feature Extraction 482

19.4.4 Classification Model 484

19.5 Results and Discussion 488

19.6 Conclusion 491

References 491

Index 495

最近チェックした商品