haebom
Daily Arxiv
전 세계에서 발간되는 인공지능 관련 논문을 정리하는 페이지 입니다.
본 페이지는 Google Gemini를 활용해 요약 정리하며, 비영리로 운영 됩니다.
논문에 대한 저작권은 저자 및 해당 기관에 있으며, 공유 시 출처만 명기하면 됩니다.
CTRL-RAG: Contrastive Likelihood Reward Based Reinforcement Learning for Context-Faithful RAG Models
Lost in Translation: How Language Re-Aligns Vision for Cross-Species Pathology
FinRetrieval: A Benchmark for Financial Data Retrieval by AI Agents
A theoretical model of dynamical grammatical gender shifting based on set-valued set function
The Spike, the Sparse and the Sink: Anatomy of Massive Activations and Attention Sinks
Towards Provably Unbiased LLM Judges via Bias-Bounded Evaluation
Dissociating Direct Access from Inference in AI Introspection
Judge Reliability Harness: Stress Testing the Reliability of LLM Judges
Legal interpretation and AI: from expert systems to argumentation and LLMs
PACE: A Personalized Adaptive Curriculum Engine for 9-1-1 Call-taker Training
Ailed: A Psyche-Driven Chess Engine with Dynamic Emotional Modulation
Building AI Coding Agents for the Terminal: Scaffolding, Harness, Context Engineering, and Lessons Learned
UniSTOK: Uniform Inductive Spatio-Temporal Kriging
WebChain: A Large-Scale Human-Annotated Dataset of Real-World Web Interaction Traces
STRUCTUREDAGENT: Planning with AND/OR Trees for Long-Horizon Web Tasks
X-RAY: Mapping LLM Reasoning Capability via Formalized and Calibrated Probes
GCAgent: Enhancing Group Chat Communication through Dialogue Agents System
Reclaiming Lost Text Layers for Source-Free Cross-Domain Few-Shot Learning
AI+HW 2035: Shaping the Next Decade
KARL: Knowledge Agents via Reinforcement Learning
MedCoRAG: Interpretable Hepatology Diagnosis via Hybrid Evidence Retrieval and Multispecialty Consensus
Bidirectional Curriculum Generation: A Multi-Agent Framework for Data-Efficient Mathematical Reasoning
Jagarin: A Three-Layer Architecture for Hibernating Personal Duty Agents on Mobile
WebFactory: Automated Compression of Foundational Language Intelligence into Grounded Web Agents
Enhancing Zero-shot Commonsense Reasoning by Integrating Visual Knowledge via Machine Imagination
The Trilingual Triad Framework: Integrating Design, AI, and Domain Knowledge in No-code AI Smart City Course
AegisUI: Behavioral Anomaly Detection for Structured User Interface Protocols in AI Agent Systems
Survive at All Costs: Exploring LLM's Risky Behaviors under Survival Pressure
S5-SHB Agent: Society 5.0 enabled Multi-model Agentic Blockchain Framework for Smart Home
Measuring the Fragility of Trust: Devising Credibility Index via Explanation Stability (CIES) for Business Decision Support Systems
BioLLMAgent: A Hybrid Framework with Enhanced Structural Interpretability for Simulating Human Decision-Making in Computational Psychiatry
Rethinking Representativeness and Diversity in Dynamic Data Selection
Retrieval-Augmented Generation with Covariate Time Series
TimeWarp: Evaluating Web Agents by Revisiting the Past
Knowledge-informed Bidding with Dual-process Control for Online Advertising
Alignment Backfire: Language-Dependent Reversal of Safety Interventions Across 16 Languages in LLM Multi-Agent Systems
EvoTool: Self-Evolving Tool-Use Policy Optimization in LLM Agents via Blame-Aware Mutation and Diversity-Aware Selection
Authorize-on-Demand: Dynamic Authorization with Legality-Aware Intellectual Property Protection for VLMs
Differentially Private Multimodal In-Context Learning
Bounded State in an Infinite Horizon: Proactive Hierarchical Memory for Ad-Hoc Recall over Streaming Dialogues
SEA-TS: Self-Evolving Agent for Autonomous Code Generation of Time Series Forecasting Algorithms
K-Gen: A Multimodal Language-Conditioned Approach for Interpretable Keypoint-Guided Trajectory Generation
Causally Robust Reward Learning from Reason-Augmented Preference Feedback
On Multi-Step Theorem Prediction via Non-Parametric Structural Priors
Design Behaviour Codes (DBCs): A Taxonomy-Driven Layered Governance Benchmark for Large Language Models
VISA: Value Injection via Shielded Adaptation for Personalized LLM Alignment
LLM-Grounded Explainability for Port Congestion Prediction via Temporal Graph Attention Networks
EchoGuard: An Agentic Framework with Knowledge-Graph Memory for Detecting Manipulative Communication in Longitudinal Dialogue
Timer-S1: A Billion-Scale Time Series Foundation Model with Serial Scaling
Breaking Contextual Inertia: Reinforcement Learning with Single-Turn Anchors for Stable Multi-Turn Interaction
MOOSEnger -- a Domain-Specific AI Agent for the MOOSE Ecosystem
Evaluating the Search Agent in a Parallel World
HiMAP-Travel: Hierarchical Multi-Agent Planning for Long-Horizon Constrained Travel
Visioning Human-Agentic AI Teaming: Continuity, Tension, and Future Research
CONE: Embeddings for Complex Numerical Data Preserving Unit and Variable Semantics
Memory as Ontology: A Constitutional Memory Architecture for Persistent Digital Citizens
Interactive Benchmarks
Solving an Open Problem in Theoretical Physics using AI-Assisted Discovery
From Offline to Periodic Adaptation for Pose-Based Shoplifting Detection in Real-world Retail Security
Model Medicine: A Clinical Framework for Understanding, Diagnosing, and Treating AI Models
Using Vision + Language Models to Predict Item Difficulty
When Agents Persuade: Propaganda Generation and Mitigation in LLMs
Towards automated data analysis: A guided framework for LLM-based risk estimation
ECG-MoE: Mixture-of-Expert Electrocardiogram Foundation Model
Self-Attribution Bias: When AI Monitors Go Easy on Themselves
Adaptive Memory Admission Control for LLM Agents
Discovering mathematical concepts through a multi-agent system
Progressive Refinement Regulation for Accelerating Diffusion Language Model Decoding
Capability Thresholds and Manufacturing Topology: How Embodied Intelligence Triggers Phase Transitions in Economic Geography
SkillNet: Create, Evaluate, and Connect AI Skills
AOI: Turning Failed Trajectories into Training Signals for Autonomous Cloud Diagnosis
ACES: Accent Subspaces for Coupling, Explanations, and Stress-Testing in Automatic Speech Recognition
Inhibitory Cross-Talk Enables Functional Lateralization in Attention-Coupled Latent Memory
Non-Invasive Reconstruction of Intracranial EEG Across the Deep Temporal Lobe from Scalp EEG based on Conditional Normalizing Flow
Perfect score on IPhO 2025 theory by Gemini agent
Physics-constrained symbolic regression for discovering closed-form equations of multimodal water retention curves from experimental data
GreenPhase: A Green Learning Approach for Earthquake Phase Picking
Neuro-Symbolic Decoding of Neural Activity
Cryo-SWAN: the Multi-Scale Wavelet-decomposition-inspired Autoencoder Network for molecular density representation of molecular volumes
Ethical and Explainable AI in Reusable MLOps Pipelines
Fragile Thoughts: How Large Language Models Handle Chain-of-Thought Perturbations
PulseLM: A Foundation Dataset and Benchmark for PPG-Text Learning
Certainty robustness: Evaluating LLM stability under self-challenging prompts
AutoHarness: improving LLM agents by automatically synthesizing a code harness
StructLens: A Structural Lens for Language Models via Maximum Spanning Trees
A benchmark for joint dialogue satisfaction, emotion recognition, and emotion state transition prediction
Controllable and explainable personality sliders for LLMs at inference time
IntPro: A Proxy Agent for Context-Aware Intent Understanding via Retrieval-conditioned Inference
Controlling Chat Style in Language Models via Single-Direction Editing
Discern Truth from Falsehood: Reducing Over-Refusal via Contrastive Refinement
Can Large Language Models Derive New Knowledge? A Dynamic Benchmark for Biological Knowledge Discovery
DIALEVAL: Automated Type-Theoretic Evaluation of LLM Instruction Following
From We to Me: Theory Informed Narrative Shift with Abductive Reasoning
Automated Concept Discovery for LLM-as-a-Judge Preference Analysis
Quantum-Inspired Self-Attention in a Large Language Model
The Influence of Iconicity in Transfer Learning for Sign Language Recognition
M-QUEST -- Meme Question-Understanding Evaluation on Semantics and Toxicity
Towards Self-Robust LLMs: Intrinsic Prompt Noise Resistance via CoIPO
How does fine-tuning improve sensorimotor representations in large language models?
Escaping the BLEU Trap: A Signal-Grounded Framework with Decoupled Semantic Guidance for EEG-to-Text Decoding
Load more
Watermark Overwriting Attack on StegaStamp algorithm
Created by
Haebom
Category
Empty
저자
I. F. Serzhenko, L. A. Khaertdinova, M. A. Pautov, A. V. Antsiferova
개요
본 논문은 NeurIPS "Erasing the invisible" 경진대회의 일환으로 개발된 StegaStamp 워터마킹 알고리즘에 대한 공격 방법을 제시한다. 이 방법은 이미지에서 워터마크를 완전히 제거하면서 이미지 품질 저하를 최소화한다.
시사점, 한계점
•
시사점:
StegaStamp 워터마킹 알고리즘의 취약성을 보여줌으로써, 워터마킹 기술의 안전성에 대한 재고를 촉구한다. 워터마킹 기법 개발에 있어 더욱 강력하고 안전한 알고리즘 개발의 필요성을 강조한다.
•
한계점:
현재 StegaStamp 알고리즘에 대한 공격에만 초점을 맞추고 있으며, 다른 워터마킹 알고리즘에 대한 일반화 가능성은 제시되지 않았다. 제안된 공격 방법의 계산 복잡도 및 효율성에 대한 자세한 분석이 부족하다.
PDF 보기
Made with Slashpage
