KI 2025: 48TH GERMAN CONFERENCE ON ARTIFICIAL INTELLIGENCE

This page contains an index consisting of author-provided keywords.

A
abstraction	Towards Observing the Effect of Abstraction on Understandability of Explanations in Answer Set Programming
academic career	How (not) to PhD Welcome
adversarial attacks	Verification of Neural Networks
AI ethics	Magic wands and digital zombies: some promises and risks of AI for digital legacy
AI Grid	Why is Networking Important in Research? A Short Introduction to the AI Grid
Anonymization	Towards Systematic Evaluation of Computer Vision Models under Data Anonymization
answer set programming	Towards Observing the Effect of Abstraction on Understandability of Explanations in Answer Set Programming
Applications	Tackling Challenges in Critical Infrastructure through Machine Learning
Applications of logic	Theory of mind - Between natural and artificial intelligence
Argumentation	Numbers Don't Lie: Hybrid Extraction and Validation of Quantitative Statements in Arguments with Semi-Structured Information
Artificial Intelligence	Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data Augmenting Systematic Literature Reviews in Information Systems: A Human-AI Collaborative Framework KI 2025 - Closing Economically-Driven AI Process for Quality Assurance: Analysis in Optics Manufacturing How (not) to PhD Why is Networking Important in Research? A Short Introduction to the AI Grid Closing, Open Discussion Round Welcome
Artificial intelligence	The origins of AI research in the Federal Republic of Germany Theory of mind - Between natural and artificial intelligence KI 2025 - Opening
attention	Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data
Automated Report Generation	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
automl	Enhancing Semi-Supervised Learning with a Meta-Feature Based Safeguard System
B
Benchmark	Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data On the Domain Robustness of Contrastive Vision-Language Models
Beta-VAE	Learning Interpretable Disentangled Concepts for Neurosymbolic Integration
Black-Box Explanations	Makrut Attacks Against Black-Box Explanations
C
Chess Agents	Caïssa AI: A Neuro-Symbolic Chess Agent for Explainable Move Suggestion and Grounded Commentary
Chest Radiography	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
CLIP: Contrastive Language-Image Pre-training	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
Code Models	Exploiting Contexts of LLM-based Code-Completion
cognitive study	Towards Observing the Effect of Abstraction on Understandability of Explanations in Answer Set Programming
Computational Geometry	Research Proposal: Runtime Verification on Spatial Objects
Computer Vision	Towards Systematic Evaluation of Computer Vision Models under Data Anonymization Verification of Neural Networks
Concept Embedding	Analyzing Deep Generative Models for Steel Microstructures
Concept-based Explanation	Toward Short and Robust Contrastive Explanations for Image Classification by Leveraging Instance Similarity and Concept Relevance
Conference closing	KI 2025 - Closing
Constraint Satisfaction and Optimization	A Hybrid Constraint-Based, Greedy, and Local Search Approach for the Transshipment Problem
Contrastive Explanation	Toward Short and Robust Contrastive Explanations for Image Classification by Leveraging Instance Similarity and Concept Relevance
Contrastive Explanation Method	Re-Evaluating the Robustness and Interpretability of the Contrastive Explanation Method for Image Classification
Contrastive Learning	Data-Efficient Multimodal Training Strategies
Cooperative Game Theory	Balanced Reciprocity for Data Sharing - Axiomatization and Mechanism Design
Coreset Selection	Data-Efficient Multimodal Training Strategies
Curriculum Learning	Data-Efficient Multimodal Training Strategies
D
DAAD	International Research Stays for Computer Scientists
Data Efficiency	Data-Efficient Multimodal Training Strategies
data leakage	Privacy Risk Assessment in Federated Learning: Extracting and Protecting Sensitive Information from Vision Language Models in Manufacturing Applications
Data Poisoning	Exploiting Contexts of LLM-based Code-Completion
data privacy	Privacy Risk Assessment in Federated Learning: Extracting and Protecting Sensitive Information from Vision Language Models in Manufacturing Applications
Data-driven emulators	Deep learning emulators for large-scale, high-resolution urban pluvial flood prediction
Decision Support Systems	Learn, Optimize, Explain: A Neuro-Symbolic Advisor for Personal Finance
Deep Generative Model	Analyzing Deep Generative Models for Steel Microstructures
Deep Learning	Deep learning emulators for large-scale, high-resolution urban pluvial flood prediction Model Efficiency Techniques in Multimodal Learning Promoting Flatness of Representation Manifolds to Improve Deep Network Training
Digital legacy	Magic wands and digital zombies: some promises and risks of AI for digital legacy
Digital Platform	Economically-Driven AI Process for Quality Assurance: Analysis in Optics Manufacturing
Disentangled Representation Learning	Learning Interpretable Disentangled Concepts for Neurosymbolic Integration
doctoral consortium	Closing, Open Discussion Round Welcome
Domain Gap	Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data
Domain Shift	On the Domain Robustness of Contrastive Vision-Language Models
DPO: Direct Preference Optimization	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
DRL	Learning Interpretable Disentangled Concepts for Neurosymbolic Integration
E
Easy Language	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Evaluation framework	ODExAI: A Comprehensive Object Detection Explainable AI Evaluation
Expert Study	Comparing the visual quality of deep generative models for steel microstructures
Explainability	Tackling Challenges in Critical Infrastructure through Machine Learning
eXplainable AI	ODExAI: A Comprehensive Object Detection Explainable AI Evaluation Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data Analyzing Deep Generative Models for Steel Microstructures
Explainable Artificial Intelligence	Toward Short and Robust Contrastive Explanations for Image Classification by Leveraging Instance Similarity and Concept Relevance Re-Evaluating the Robustness and Interpretability of the Contrastive Explanation Method for Image Classification Economically-Driven AI Process for Quality Assurance: Analysis in Optics Manufacturing Explainable Artificial Intelligence for Multivariate Sensor Data: Towards Transparency and Correctness in Model Explanations
explainable neural network	Verification of Neural Networks
Explainable Reasoning	Caïssa AI: A Neuro-Symbolic Chess Agent for Explainable Move Suggestion and Grounded Commentary
Explanation Complexity	Toward Short and Robust Contrastive Explanations for Image Classification by Leveraging Instance Similarity and Concept Relevance
Explanation Robustness	Toward Short and Robust Contrastive Explanations for Image Classification by Leveraging Instance Similarity and Concept Relevance
explanations	Towards Observing the Effect of Abstraction on Understandability of Explanations in Answer Set Programming
F
Fairness	Balanced Reciprocity for Data Sharing - Axiomatization and Mechanism Design
Feature Selection	Unsupervised Selection of Features by their Resilience to the Curse of Dimensionality
feature visualization	Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data
Federated Learning	Privacy Risk Assessment in Federated Learning: Extracting and Protecting Sensitive Information from Vision Language Models in Manufacturing Applications
Fine-Tuning	Intermediate-Task Transfer Learning for Bioacoustic Data
Formal Methods	Research Proposal: Runtime Verification on Spatial Objects
Foundation Models	Promoting Flatness of Representation Manifolds to Improve Deep Network Training
G
generative AI	Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data
Generative Artificial Intelligence	Comparing the visual quality of deep generative models for steel microstructures
German conference on artificial intelligence	KI 2025 - Opening
German conference on artificial intelligence	KI 2025 - Closing
Germany	The origins of AI research in the Federal Republic of Germany
Governmental Texts	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Graph Learning	Unsupervised Selection of Features by their Resilience to the Curse of Dimensionality
Graph Neural Networks	Efficient Graph-Based Neural Architectures for Multimodal Learning
H
History	The origins of AI research in the Federal Republic of Germany
Human-AI Collaboration	Augmenting Systematic Literature Reviews in Information Systems: A Human-AI Collaborative Framework
Hybrid Datasets	Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data
I
Image Classification	Re-Evaluating the Robustness and Interpretability of the Contrastive Explanation Method for Image Classification
Image-to-image regression	Deep learning emulators for large-scale, high-resolution urban pluvial flood prediction
In-context learning	Re-examining learning linear functions in context
Inductive Logic Programming	Learning Interpretable Disentangled Concepts for Neurosymbolic Integration
International research stays	International Research Stays for Computer Scientists
Interpretability	Re-Evaluating the Robustness and Interpretability of the Contrastive Explanation Method for Image Classification
interpretable machine learning	Explainable Artificial Intelligence for Multivariate Sensor Data: Towards Transparency and Correctness in Model Explanations
Intralingual Translation	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Intrinsic Dimensionality	Unsupervised Selection of Features by their Resilience to the Curse of Dimensionality
Intrinsically Interpretable Model	Analyzing Deep Generative Models for Steel Microstructures
L
Language Models	Accessible Language Simplification: Large Language Models for Generating Easy German
Large Language Models	Augmenting Systematic Literature Reviews in Information Systems: A Human-AI Collaborative Framework Numbers Don't Lie: Hybrid Extraction and Validation of Quantitative Statements in Arguments with Semi-Structured Information Exploiting Contexts of LLM-based Code-Completion
Leichte Sprache	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Length Generalization	Positional Overload: Positional Debiasing and Context Window Extension for Large Language Models using Set Encoding
Lifted Bayesian Networks	Probabilities of the Third Type: Statistical Relational Learning and Reasoning with Relative Frequencies (Extended Abstract)
LIME	Makrut Attacks Against Black-Box Explanations
Linear Temporal Logic	Research Proposal: Runtime Verification on Spatial Objects
LLM	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
LLMs	Re-examining learning linear functions in context Positional Overload: Positional Debiasing and Context Window Extension for Large Language Models using Set Encoding
LLMs: Large Language Models	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
M
Machine Learning	Enhancing Semi-Supervised Learning with a Meta-Feature Based Safeguard System Model Efficiency Techniques in Multimodal Learning
Machine Learning	Tackling Challenges in Critical Infrastructure through Machine Learning
Machine Translation	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Manifold Hypothesis	Promoting Flatness of Representation Manifolds to Improve Deep Network Training
manufacturing data	Privacy Risk Assessment in Federated Learning: Extracting and Protecting Sensitive Information from Vision Language Models in Manufacturing Applications
Mechanism Design	Balanced Reciprocity for Data Sharing - Axiomatization and Mechanism Design
meta-features	Enhancing Semi-Supervised Learning with a Meta-Feature Based Safeguard System
Mixed Training	Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data
Model and Data Sharing	Balanced Reciprocity for Data Sharing - Axiomatization and Mechanism Design
Model efficiency	Model Efficiency Techniques in Multimodal Learning
Model Inversion Attack	Privacy Risk Assessment in Federated Learning: Extracting and Protecting Sensitive Information from Vision Language Models in Manufacturing Applications
Model Memorization	Privacy Risk Assessment in Federated Learning: Extracting and Protecting Sensitive Information from Vision Language Models in Manufacturing Applications
Multi-agent systems	Theory of mind - Between natural and artificial intelligence
Multimodal AI: Multimodal Artificial Intelligence	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
Multimodal Learning	Efficient Graph-Based Neural Architectures for Multimodal Learning Model Efficiency Techniques in Multimodal Learning Data-Efficient Multimodal Training Strategies
multivariate time series	Explainable Artificial Intelligence for Multivariate Sensor Data: Towards Transparency and Correctness in Model Explanations
N
networking	Why is Networking Important in Research? A Short Introduction to the AI Grid
neural network verification	Verification of Neural Networks
Neuro-Symbolic AI	Caïssa AI: A Neuro-Symbolic Chess Agent for Explainable Move Suggestion and Grounded Commentary Learn, Optimize, Explain: A Neuro-Symbolic Advisor for Personal Finance
O
Object Detection	ODExAI: A Comprehensive Object Detection Explainable AI Evaluation Towards Systematic Evaluation of Computer Vision Models under Data Anonymization Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data XAIRob --- An Explainable-AI-Based Relative Robustness Measure for Object Detection
Opening	KI 2025 - Opening
Optical Microscopy	Comparing the visual quality of deep generative models for steel microstructures
Optics Manufacturing	Economically-Driven AI Process for Quality Assurance: Analysis in Optics Manufacturing
P
Passive Acoustic Monitoring	Intermediate-Task Transfer Learning for Bioacoustic Data
Personal Finance	Learn, Optimize, Explain: A Neuro-Symbolic Advisor for Personal Finance
Plain Language	Accessible Language Simplification: Large Language Models for Generating Easy German
Planning	A Hybrid Constraint-Based, Greedy, and Local Search Approach for the Transshipment Problem
point clouds	XAIRob --- An Explainable-AI-Based Relative Robustness Measure for Object Detection
Positional Debiasing	Positional Overload: Positional Debiasing and Context Window Extension for Large Language Models using Set Encoding
Positional Embeddings	Positional Overload: Positional Debiasing and Context Window Extension for Large Language Models using Set Encoding
Probability Logic	Probabilities of the Third Type: Statistical Relational Learning and Reasoning with Relative Frequencies (Extended Abstract)
Q
Quality Assurance	Economically-Driven AI Process for Quality Assurance: Analysis in Optics Manufacturing
Quantitative Statements	Numbers Don't Lie: Hybrid Extraction and Validation of Quantitative Statements in Arguments with Semi-Structured Information
Question Answering	Accessible Language Simplification: Large Language Models for Generating Easy German
R
RAG: Retrieval-Augmented Generation	AI-Driven Chest Radiography Report Generation: Integrating LLMs, CLIP, Tree-of-Thoughts, Multimodal Retrieval-Augmented Generation, Classification and Direct Preference Optimization
Representation Learning	Promoting Flatness of Representation Manifolds to Improve Deep Network Training
research agenda	How (not) to PhD
Research Augmentation	Augmenting Systematic Literature Reviews in Information Systems: A Human-AI Collaborative Framework
Resource optimization	Model Efficiency Techniques in Multimodal Learning
Retrieval Augmented Generation	Accessible Language Simplification: Large Language Models for Generating Easy German
Retrieval-Augmented Generation	Numbers Don't Lie: Hybrid Extraction and Validation of Quantitative Statements in Arguments with Semi-Structured Information
Reward Allocation	Balanced Reciprocity for Data Sharing - Axiomatization and Mechanism Design
RGB images	XAIRob --- An Explainable-AI-Based Relative Robustness Measure for Object Detection
Robustness	Re-Evaluating the Robustness and Interpretability of the Contrastive Explanation Method for Image Classification XAIRob --- An Explainable-AI-Based Relative Robustness Measure for Object Detection
Robustness Evaluation	On the Domain Robustness of Contrastive Vision-Language Models
Routing	A Hybrid Constraint-Based, Greedy, and Local Search Approach for the Transshipment Problem
Runtime Verification	Research Proposal: Runtime Verification on Spatial Objects
S
Scheduling	A Hybrid Constraint-Based, Greedy, and Local Search Approach for the Transshipment Problem
Scholarships	International Research Stays for Computer Scientists
Search and Optimization	A Hybrid Constraint-Based, Greedy, and Local Search Approach for the Transshipment Problem
Security of XAI	Makrut Attacks Against Black-Box Explanations
semi-supervised learning	Enhancing Semi-Supervised Learning with a Meta-Feature Based Safeguard System
sensor data	Explainable Artificial Intelligence for Multivariate Sensor Data: Towards Transparency and Correctness in Model Explanations
Spatial Logic	Research Proposal: Runtime Verification on Spatial Objects
Spatio-Temporal Graphs	Efficient Graph-Based Neural Architectures for Multimodal Learning
Statistical Relational Learning	Probabilities of the Third Type: Statistical Relational Learning and Reasoning with Relative Frequencies (Extended Abstract)
Steel microstructure	Comparing the visual quality of deep generative models for steel microstructures
Steel Microstructures	Analyzing Deep Generative Models for Steel Microstructures
Subsurface Damage	Economically-Driven AI Process for Quality Assurance: Analysis in Optics Manufacturing
Synthetic Data	Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data
Systematic Literature Review	Augmenting Systematic Literature Reviews in Information Systems: A Human-AI Collaborative Framework
T
Text Simplification	LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages
Thanabots	Magic wands and digital zombies: some promises and risks of AI for digital legacy
Time-series Analysis	Efficient Graph-Based Neural Architectures for Multimodal Learning
Transfer Learning	Intermediate-Task Transfer Learning for Bioacoustic Data Visualizing and Interpreting Neural Network Focus Regions: A Comparative Study of Vision Transformers on Synthetic and Real Data
Transformers	Re-examining learning linear functions in context Efficient Graph-Based Neural Architectures for Multimodal Learning
Transshipment Problem	A Hybrid Constraint-Based, Greedy, and Local Search Approach for the Transshipment Problem
U
Unsupervised	Learning Interpretable Disentangled Concepts for Neurosymbolic Integration
Urban flood prediction	Deep learning emulators for large-scale, high-resolution urban pluvial flood prediction
V
Validation	Numbers Don't Lie: Hybrid Extraction and Validation of Quantitative Statements in Arguments with Semi-Structured Information
Vision-Language Models	On the Domain Robustness of Contrastive Vision-Language Models
Visual Concepts	Analyzing Deep Generative Models for Steel Microstructures
X
x-Ai	XAIRob --- An Explainable-AI-Based Relative Robustness Measure for Object Detection
XAI	Explainable Artificial Intelligence for Multivariate Sensor Data: Towards Transparency and Correctness in Model Explanations
Y
young researchers	Closing, Open Discussion Round

Disclaimer | Powered by EasyChair Smart Program