전체 논문 인덱스 — arXiv 96편
이 볼트가 인용한 arXiv 논문 96편을 챕터별로 모았다. ID를 누르면 arxiv.org 원문으로, “인용 노트”의 위키링크를 누르면 그 논문을 다루는 볼트 노트로 넘어간다. 이 가운데 22편은 2026-06-20 보강에서 추가한 2026년 신작이다.
챕터 배정 기준은 2026년 신작 22편은 주제, 나머지는 인용 노트다. 한 논문이 여러 챕터에서 인용되면 대표 챕터에만 싣고, 인용 노트 열에는 관련 노트를 모두 표기했다.
00 개요 (32편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 1706.03762 | Attention Is All You Need | 00_03-Transformer-어텐션-KV캐시 |
| 2209.11895 | In-context Learning and Induction Heads | — |
| 2212.08073 | Constitutional AI: Harmlessness from AI Feedback | — |
| 2305.14325 | Improving Factuality and Reasoning in Language Models through Multiagent Debate | — |
| 2311.09198 | Never Lost in the Middle: Mastering Long-Context Question Answering with Position-Agnostic Decompositional Training | — |
| 2403.04797 | Found in the Middle: How Language Models Use Long Contexts Better via Plug-and-Play Positional Encoding | — |
| 2410.02694 | HELMET: How to Evaluate Long-Context Language Models Effectively and Thoroughly | — |
| 2410.10813 | LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory | — |
| 2411.03538 | Long Context RAG Performance of Large Language Models | 00_04-컨텍스트-로트-Context-Rot · 06_10-RAG-vs-롱컨텍스트-논쟁 |
| 2411.15399 | Less is More: Optimizing Function Calling for LLM Execution on Edge Devices | 00_06-5대-실패모드-분류 · 00_실패모드-진단-퀴즈 · 03_03-소형모델-취약성 |
| 2412.19442 | A Survey on Large Language Model Acceleration based on KV Cache Management | 00_03-Transformer-어텐션-KV캐시 |
| 2501.06322 | Multi-Agent Collaboration Mechanisms: A Survey of LLMs | — |
| 2502.08177 | SycEval: Evaluating LLM Sycophancy | — |
| 2502.19130 | Voting or Consensus? Decision-Making in Multi-Agent Debate | — |
| 2504.00180 | Contradiction Detection in RAG Systems: Evaluating LLMs as Context Validators for Improved Information Consistency | — |
| 2505.02077 | Open Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents | — |
| 2505.02279 | A survey of agent interoperability protocols: Model Context Protocol (MCP), Agent Communication Protocol (ACP), Agent-to-Agent Protocol (A2A), and Agent Network Protocol (ANP) | — |
| 2505.06120 | LLMs Get Lost In Multi-Turn Conversation | 00_06-5대-실패모드-분류 · 00_실패모드-진단-퀴즈 · 00_핵심용어집 |
| 2505.15392 | Understanding the Anchoring Effect of LLM with Synthetic Data: Existence, Mechanism, and Potential Mitigations | — |
| 2507.06261 | Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities | 00_06-5대-실패모드-분류 · 00_실패모드-진단-퀴즈 · 04_09-에이전트-취약성-왜-더-심각한가 |
| 2510.00615 | ACON: Optimizing Context Compression for Long-horizon LLM Agents | — |
| 2510.26493 | Context Engineering 2.0: The Context of Context Engineering | 00_01-용어-탄생과-계보 |
| 2601.02371 | Permission Manifests for Web Agents | — |
| 2602.22724 | AgentSentry: Mitigating Indirect Prompt Injection in LLM Agents via Temporal Causal Diagnostics and Context Purification | — |
| 2603.03308 | Old Habits Die Hard: How Conversational History Geometrically Traps LLMs | — |
| 2603.10123 | Lost in the Middle at Birth: An Exact Theory of Transformer Position Bias | — |
| 2603.20397 | KV Cache Optimization Strategies for Scalable and Efficient LLM Inference | 00_03-Transformer-어텐션-KV캐시 |
| 2604.01664 | ContextBudget: Budget-Aware Context Management for Long-Horizon Search Agents | 00_01-용어-탄생과-계보 · 00_04-컨텍스트-로트-Context-Rot · 00_05-에이전트-컨텍스트-누적 · 00_07-4대-전략-Write-Select-Compress-Isolate |
| 2604.07007 | AgentCity: Constitutional Governance for Autonomous Agent Economies via Separation of Power | — |
| 2604.07911 | Dynamic Attentional Context Scoping: Agent-Triggered Focus Sessions for Isolated Per-Agent Steering in Multi-Agent LLM Orchestration | — |
| 2604.16339 | Semantic Consensus: Process-Aware Conflict Detection and Resolution for Enterprise Multi-Agent LLM Systems | — |
| 2606.20245 | Navigating Unreliable Parametric and Contextual Knowledge: Explicit Knowledge Conflict Resolution for LLM Inference | — |
01 오염 (16편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2402.07867 | PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models | 01_04-우발-vs-적대-비교 · 01_07-paper-memorygraft · 01_08-적대적-오염-위협지형 · 01_10-논쟁점-미해결과제 |
| 2407.12784 | AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases | 01_04-우발-vs-적대-비교 · 01_07-paper-memorygraft · 01_08-적대적-오염-위협지형 · 01_10-논쟁점-미해결과제 |
| 2503.18813 | Defeating Prompt Injections by Design | 01_00_MOC · 01_04-우발-vs-적대-비교 · 01_05-paper-agentpoison · 01_07-paper-memorygraft · 01_08-적대적-오염-위협지형 · 01_09-처방-방어전략 · 01_11-paper-CaMeL-방어 · 07_08-처방-컨텍스트는-슬롯계약-4계층-모델 |
| 2506.23260 | From Prompt Injections to Protocol Exploits: Threats in LLM-Powered AI Agents Workflows | 01_05-paper-agentpoison · 01_08-적대적-오염-위협지형 |
| 2509.10540 | EchoLeak: The First Real-World Zero-Click Prompt Injection Exploit in a Production LLM System | 01_12-사례-산업-실사고 |
| 2512.16962 | MemoryGraft: Persistent Compromise of LLM Agents via Poisoned Experience Retrieval | 01_04-우발-vs-적대-비교 · 01_07-paper-memorygraft |
| 2601.05504 | Memory Poisoning Attack and Defense on Memory Based LLM-Agents | 01_09-처방-방어전략 · 01_10-논쟁점-미해결과제 |
| 2601.05755 | VIGIL: Defending LLM Agents Against Tool Stream Injection via Verify-Before-Commit | 09_09-해설-오염-깊이읽기 |
| 2601.10923 | Hidden-in-Plain-Text: A Benchmark for Social-Web Indirect Prompt Injection in RAG | 09_09-해설-오염-깊이읽기 |
| 2602.04711 | Addressing Corpus Knowledge Poisoning Attacks on RAG Using Sparse Attention | 09_09-해설-오염-깊이읽기 |
| 2602.15654 | Zombie Agents: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing Injections | 09_09-해설-오염-깊이읽기 |
| 2602.21447 | Adversarial Intent is a Latent Variable: Stateful Trust Inference for Securing Multimodal Agentic RAG | 09_09-해설-오염-깊이읽기 |
| 2603.11768 | Governing Evolving Memory in LLM Agents: Risks, Mechanisms, and the Stability and Safety Governed Memory (SSGM) Framework | 01_02-메커니즘-자기강화-루프 · 01_04-우발-vs-적대-비교 · 01_09-처방-방어전략 · 01_10-논쟁점-미해결과제 |
| 2604.07536 | TRUSTDESC: Preventing Tool Poisoning in LLM Applications via Trusted Description Generation | 09_09-해설-오염-깊이읽기 |
| 2604.08304 | Securing Retrieval-Augmented Generation: A Taxonomy of Attacks, Defenses, and Future Directions | 09_09-해설-오염-깊이읽기 |
| 2606.04329 | From Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agents | 09_09-해설-오염-깊이읽기 |
02 산만 (5편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2306.15595 | Extending Context Window of Large Language Models via Positional Interpolation | 02_03-메커니즘-위치-편향-u자형 |
| 2309.00071 | YaRN: Efficient Context Window Extension of Large Language Models | 02_03-메커니즘-위치-편향-u자형 |
| 2309.17453 | Efficient Streaming Language Models with Attention Sinks | 02_03-메커니즘-위치-편향-u자형 · 05_03-메커니즘-왜-발생하는가 |
| 2410.18745 | Why Does the Effective Context Length of LLMs Fall Short? | 02_02-메커니즘-어텐션-희석 · 02_03-메커니즘-위치-편향-u자형 · 02_04-개념-산만-임계점 · 02_10-논쟁-산만-vs-훈련부족-vs-아키텍처 · 02_11-처방-컨텍스트-산만-완화-체크리스트 |
| 2510.05381 | Context Length Alone Hurts LLM Performance Despite Perfect Retrieval | 02_01-정의-컨텍스트-산만 · 02_02-메커니즘-어텐션-희석 · 02_04-개념-산만-임계점 · 02_10-논쟁-산만-vs-훈련부족-vs-아키텍처 · 02_11-처방-컨텍스트-산만-완화-체크리스트 · 05_01-정의-컨텍스트-부패 · 05_02-분류-다른-실패모드와의-관계 · 05_03-메커니즘-왜-발생하는가 · 05_04-paper-context-rot-chroma · 05_07-처방-완화-체크리스트 |
03 혼란 (9편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2505.03275 | RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation | 03_05-paper-rag-mcp |
| 2505.06416 | ScaleMCP: Dynamic and Auto-Synchronizing Model Context Protocol Tools for LLM Agents | 03_05-paper-rag-mcp |
| 2512.24565 | MCPAgentBench: A Real-world Task Benchmark for Evaluating LLM Agent MCP Tool Use | 09_11-해설-혼란-깊이읽기 |
| 2601.05214 | Internal Representations as Indicators of Hallucinations in Agent Tool Selection | 03_08-해결전략-less-is-more |
| 2602.00933 | MCP-Atlas: A Large-Scale Benchmark for Tool-Use Competency with Real MCP Servers | 09_11-해설-혼란-깊이읽기 |
| 2602.20426 | Learning to Rewrite Tool Descriptions for Reliable LLM-Agent Tool Use | 03_08-해결전략-less-is-more |
| 2605.24660 | How Many Tools Should an LLM Agent See? A Chance-Corrected Answer | 03_03-소형모델-취약성 · 03_06-paper-cmtf-bor |
| 2606.06284 | ToolChoiceConfusion: Causal Minimal Tool Filtering for Reliable LLM Agents | 03_06-paper-cmtf-bor |
| 2606.10209 | Less Context, Better Agents: Efficient Context Engineering for Long-Horizon Tool-Using LLM Agents | 03_07-paper-less-context-better-agents |
04 충돌 (10편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2511.04694 | Reasoning Up the Instruction Ladder for Controllable Language Models | 04_09-에이전트-취약성-왜-더-심각한가 |
| 2602.04288 | Contextual Drag: How Errors in the Context Affect LLM Reasoning | 04_09-에이전트-취약성-왜-더-심각한가 |
| 2603.00024 | Personalization Increases Affective Alignment but Has Role-Dependent Effects on Epistemic Independence in LLMs | 09_12-해설-충돌-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2603.10521 | IH-Challenge: A Training Dataset to Improve Instruction Hierarchy on Frontier LLMs | 09_12-해설-충돌-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2603.12123 | Cross-Context Review: Improving LLM Output Quality by Separating Production and Review Sessions | 04_09-에이전트-취약성-왜-더-심각한가 |
| 2603.16152 | HIPO: Instruction Hierarchy via Constrained Reinforcement Learning | 09_12-해설-충돌-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2604.09075 | Hierarchical Alignment: Enforcing Hierarchical Instruction-Following in LLMs through Logical Consistency | 09_12-해설-충돌-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2604.09443 | Many-Tier Instruction Hierarchy in LLM Agents | 04_09-에이전트-취약성-왜-더-심각한가 |
| 2605.27784 | Diagnosing Live Within-Policy Instruction Conflicts in LLM Agents with Witnessed Resolution Profiles | 09_12-해설-충돌-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2606.10860 | Training LLMs to Enforce Multi-Level Instruction Hierarchies via Gravity-Weighted Direct Preference Optimization | 09_12-해설-충돌-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
05 부패 (9편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2104.09864 | RoFormer: Enhanced Transformer with Rotary Position Embedding | 05_03-메커니즘-왜-발생하는가 |
| 2302.00093 | Large Language Models Can Be Easily Distracted by Irrelevant Context | 05_01-정의-컨텍스트-부패 · 05_02-분류-다른-실패모드와의-관계 · 05_04-paper-context-rot-chroma · 05_07-처방-완화-체크리스트 |
| 2307.03172 | Lost in the Middle: How Language Models Use Long Contexts | 05_02-분류-다른-실패모드와의-관계 · 05_03-메커니즘-왜-발생하는가 · 05_07-처방-완화-체크리스트 |
| 2404.06654 | RULER: What’s the Real Context Size of Your Long-Context Language Models? | 02_10-논쟁-산만-vs-훈련부족-vs-아키텍처 · 05_01-정의-컨텍스트-부패 · 05_02-분류-다른-실패모드와의-관계 · 05_04-paper-context-rot-chroma |
| 2502.05167 | NoLiMa: Long-Context Evaluation Beyond Literal Matching | 02_10-논쟁-산만-vs-훈련부족-vs-아키텍처 · 05_01-정의-컨텍스트-부패 · 05_02-분류-다른-실패모드와의-관계 · 05_03-메커니즘-왜-발생하는가 · 05_04-paper-context-rot-chroma |
| 2506.11440 | AbsenceBench: Language Models Can’t Tell What’s Missing | 05_01-정의-컨텍스트-부패 · 05_03-메커니즘-왜-발생하는가 |
| 2601.11564 | Context Discipline and Performance Correlation: Analyzing LLM Performance and Quality Degradation Under Varying Context Lengths | 09_10-해설-산만-깊이읽기 · 09_13-해설-부패-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2601.15300 | Intelligence Degradation in Long-Context LLMs: Critical Threshold Determination via Natural Length Distribution Analysis | 09_13-해설-부패-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
| 2605.12366 | Classifier Context Rot: Monitor Performance Degrades with Context Length | 09_13-해설-부패-깊이읽기 · 09_14-2026-최신동향-5대모드-업데이트 |
06 해결전략 (10편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2304.08467 | Learning to Compress Prompts with Gist Tokens | 06_03-Compress전략-요약-프루닝-KV캐시최적화 |
| 2307.13854 | WebArena: A Realistic Web Environment for Building Autonomous Agents | 06_11-에이전트-벤치마크-tau-swe-gaia |
| 2310.05736 | LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models | 06_03-Compress전략-요약-프루닝-KV캐시최적화 |
| 2310.06770 | SWE-bench: Can Language Models Resolve Real-World GitHub Issues? | 06_11-에이전트-벤치마크-tau-swe-gaia |
| 2310.08560 | MemGPT: Towards LLMs as Operating Systems | 06_01-Write전략-스크래치패드-파일시스템-크로스세션 · 06_08-MemGPT-Letta-OS형-롱텀메모리 · 06_09-실패모드-전략-매핑표 |
| 2311.12983 | GAIA: a benchmark for General AI Assistants | 06_11-에이전트-벤치마크-tau-swe-gaia |
| 2406.12045 | $τ$-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains | 06_11-에이전트-벤치마크-tau-swe-gaia |
| 2407.16833 | Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach | 06_10-RAG-vs-롱컨텍스트-논쟁 |
| 2506.07982 | $τ^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment | 06_11-에이전트-벤치마크-tau-swe-gaia |
| 2507.13334 | A Survey of Context Engineering for Large Language Models | 06_09-실패모드-전략-매핑표 |
07 거버넌스 (5편)
| arXiv | 제목 | 인용 노트 |
|---|---|---|
| 2602.17913 | From Lossy to Verified: A Provenance-Aware Tiered Memory for Agents | 07_11-provenance-운영표준-C2PA-VC |
| 2603.07670 | Memory for Autonomous LLM Agents:Mechanisms, Evaluation, and Emerging Frontiers | 09_14-2026-최신동향-5대모드-업데이트 |
| 2603.09619 | Context Engineering: From Prompts to Corporate Multi-Agent Architecture | 09_14-2026-최신동향-5대모드-업데이트 |
| 2603.18043 | The Provenance Paradox in Multi-Agent LLM Routing: Delegation Contracts and Attested Identity in LDP | 07_06-provenance와-신뢰등급-프로베넌스-역설 |
| 2604.16548 | A Survey on Long-Term Memory Security in LLM Agents: Attacks, Defenses, and Governance Across the Memory Lifecycle | 09_14-2026-최신동향-5대모드-업데이트 |
관련
- sources — 분류별 마스터 서지(블로그·공식문서 포함)
- papers-2026-신규 · articles-2026 · tools-2026 — 2026 보강 출처