Seminar | Neuro-Mechanistic Modeling

Formal Methods for AI Alignment

Wed, 01 Apr 2026 00:00:00 +0000

Team

Kevin Baum, Lisa Dargasz, Felix Jahn, Yannic Muskalla, Andre Steingrüber

Content

AI alignment is, broadly speaking, the task of ensuring that AI systems act in accordance with human values and goals (though what exactly this means remains contested). Many prevalent alignment approaches, such as RLHF and Constitutional AI, suffer from limited interpretability and lack formal guarantees. Bridging these approaches with formal frameworks, both computational and normative, seems tempting, as it promises more rigorous and interpretable alignment techniques.

In this seminar, we explored a rich repertoire of papers: foundational work on the current state of alignment methods; prima facie suitable formal approaches from computer science (including temporal logics and reward machines) and philosophy (such as deontic logics and theories of reasons); and neuro-symbolic architectures that attempt to combine the strengths of learned systems with formal methods. Participants discussed these different strands of research with one another and with members of our interdisciplinary research group.

Requirements

Genuine interest in questions related to AI alignment and AI safety
Familiarity with some of the following areas: agentic AI, reinforcement learning, formal logic and formal methods, practical reasoning, and practical philosophy
Ability and willingness to read and critically discuss English-language research papers
Openness to interdisciplinary perspectives and active, engaged participation in seminar discussions

Topics

Block A: State-of-the-Art AI Alignment

A1, RLHF basics: Deep Reinforcement Learning from Human Preferences; Training Language Models to Follow Instructions with Human Feedback
A2, RLHF-based alignment and Constitutional AI: Constitutional AI: Harmlessness from AI Feedback; Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
A3, LLM guardrails: Symbolic Guardrails for Domain-Specific Agents: Stronger Safety and Security Guarantees Without Sacrificing Utility; AgentSpec: Customizable Runtime Enforcement for Safe and Reliable LLM Agents

Block B: Normative Logics

B1, Normative systems: A Temporal Logic of Normative Systems; On social laws for artificial agent societies: off-line design
B2, Logic of reasons: Piecemeal Knowledge Acquisition for Computational Normative Reasoning; Precedent and Open Texture
B3, Deontic temporal logic: Strategic Deontic Temporal Logic as a Reduction to ATL, with an Application to Chisholm’s Scenario; Deontic Temporal Logic for Formal Verification of AI Ethics
B4, Logic and verification: Formal verification of ethical choices in autonomous systems; Specifying Agent Ethics (Blue Sky Ideas)

Block C: Logic-Based RL

C1, RL with reward machines: Synthesising Reward Machines for Cooperative Multi-Agent Reinforcement Learning; Reward Machines: Exploiting Reward Function Structure in Reinforcement Learning
C2, RL with temporal logics: Reinforcement Learning Based Temporal Logic Control with Maximum Probabilistic Satisfaction; Temporal Logic Guided Safe Reinforcement Learning Using Control Barrier Functions
C3, Ethical RL agents: Enforcing Ethical Goals Over Reinforcement Learning Policies; Learning Normative Behaviour Through Automated Theorem Proving
C4, Shielding: Safe Reinforcement Learning via Shielding; Safe Reinforcement Learning Using Probabilistic Shields

Block D: Neuro-Symbolic Architectures

D1, Neurosymbolic AI: Neurosymbolic AI: The 3rd Wave; The Third AI Summer: AAAI Robert S. Engelmore Memorial Lecture
D2, World models: Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems; The Scientist AI: Safe by Design, by Not Desiring
D3, Neurosymbolic alignment: Hybrid Approaches for Moral Value Alignment in AI Agents
D4, GRACE: Breaking Up with Normatively Monolithic Agency with GRACE: A Reason-Based Neuro-Symbolic Architecture for Safe and Ethical AI Alignment

Organization

All information and materials are available in the CMS.

Special Topics in Deep Reinforcement Learning

Wed, 01 Apr 2026 00:00:00 +0000

Team

Prof. Dr. Verena Wolf, Timo P. Gros (lecturers)
Magnus Cunow (organization)
Julius Gabelmann, Joshua Meyer, Nicola J. Müller (assistants)

Content

Reinforcement Learning (RL) is a popular sub-discipline of Machine Learning for problems that require strategies to solve complex tasks such as board games, scheduling problems, or other discrete optimization problems. Deep Reinforcement Learning (DRL) is the combination of Reinforcement Learning with Neural Networks.

In this seminar, we first covered the most important algorithms of Deep Reinforcement Learning. Afterwards, we looked at special topics that our group is conducting research on: exploration techniques, agentic RL, and AI planning.

Format

After a kick-off in April, each participant was assigned a paper and presented it in one of two seminar blocks in June, wrote a summary paper about it, and worked on a practical project in a team. The seminar ended with a closing session on the project results, with prizes, and short presentations of the group’s current research for students interested in a Master’s thesis or a HiWi position.

Topics

Deep Reinforcement Learning

Value-based DRL Evolution (core: Rainbow)
Actor-Critic: Structure & Estimation (core: A2C/A3C)
Policy Optimization Stability (core: TRPO & PPO)
Continuous Control in DRL (core: DDPG & SAC)

Planning & RL

Exploration

Agentic RL

Organization

All information and materials are available in the CMS.

Spatiotemporal Models and Inference

Tue, 01 Apr 2025 00:00:00 +0000

Team

Dr. Jonas Wahl (lecturer)
Felix Jahn, Pascal Schumann (assistants)

Content

In many scientific fields, data is not just tied to a single location or a single moment in time: it evolves over both space and time. Whether tracking the spread of infectious diseases, predicting climate patterns, modeling traffic flows, or analyzing environmental pollution, understanding how variables change across both dimensions is crucial. This is where spatiotemporal modeling comes in.

Spatiotemporal models aim to capture dependencies and relationships in data that vary over time and across geographical locations. Unlike purely spatial or purely temporal models, these approaches integrate both dimensions, allowing us to uncover trends, make predictions, and quantify uncertainties in dynamic systems. Spatiotemporal models can be of different types:

Deterministic models, based on physical laws or equations (e.g., fluid dynamics for weather modeling).
Statistical models, which use probability distributions to describe uncertainties (e.g., Gaussian processes, Bayesian hierarchical models).
Machine learning models, which learn patterns from large datasets using neural networks, random forests, or deep learning approaches.

The latter two were the main focus of this seminar. Applications of spatiotemporal modeling are found across the sciences, from forecasting the spread of diseases and analyzing temperature changes over decades to optimizing transportation networks and tracking animal migration or deforestation. The seminar covered the theoretical underpinnings of spatiotemporal statistics, Bayesian approaches and hierarchical models, and machine learning techniques for complex spatiotemporal data.

Format

A block seminar for up to 10 participants, with a kick-off in June. Each participant gave a 60-minute presentation. Grading: presentation (50%), participation in the talks of fellow participants and the discussion (25%), and a final report on the topic (25%).

Prerequisites: successful participation in a machine or deep learning course, or a background in statistical modeling, for instance from a course on statistics or stochastic processes.

Topics

Statistics for Spatiotemporal Data, Chapter 2: Statistical preliminaries
Statistics for Spatiotemporal Data, Chapter 3: Fundamentals of Temporal Processes
Statistics for Spatiotemporal Data, Chapter 4: Fundamentals of Spatial Random Processes
Introduction to Gaussian processes (lecture notes here and here, and the book)
Coding tutorial on spatial Gaussian processes (based on this blog post and this one)
Introduction to point processes (lecture notes here, slides here and here)
Coding tutorial on point processes (with the tick package, this simulator, these Jupyter notebooks, or spatial point patterns in R)
Scalable spatiotemporal prediction with Bayesian neural fields, with a coding tutorial on Bayesian neural fields
A Bayesian machine-learning approach for spatiotemporal prediction of COVID-19 cases
Spatial-temporal graph neural networks for groundwater data

Organization

All information and materials are available in the CMS.

Hands-On Graph Neural Networks

Tue, 01 Oct 2024 00:00:00 +0000

Lecturers

Joschka Groß, M.Sc.
Felix Jahn, M.Sc.
Prof. Dr. Verena Wolf

The Hands-on GNNs seminar covered the basics of graph neural networks and their applications in various domains. The seminar was held in the winter semester 2024. Participants read and presented papers on the topic and summarized their learnings in easy-to-follow-along talktorials. These are Jupyter notebooks that guide the reader through select topics concerning the fundamentals of GNNs and more advanced topics.

Organization

For all organizational matters, please first consult the CMS or contact Joschka Groß.

Final talktorials

Coming soon.

Neuro-symbolic General Policy Learning

Tue, 01 Oct 2024 00:00:00 +0000

Supervisors

Timo P. Gros, Jörg Hoffmann, Daniel Höller, Nicola J. Müller and Prof. Dr. Verena Wolf

Content

Symbolic AI relies on explicit, human-readable symbols and rules to represent knowledge and solve problems. This approach contrasts with modern, data-driven methods like machine learning, which rely on statistical patterns rather than explicit rules. Recently, both approaches have been combined, leading to the development of neuro-symbolic methods.

For instance, heuristic search, as a very successful part of symbolic AI, must be applied individually to each instance of a domain or problem. In contrast, neuro-symbolic general policy learning focuses on neural network-based policies that can solve all instances within a specific domain, combining the benefits of symbolic and data-driven AI.

In this seminar, we explored the emerging trend of learning neuro-symbolic general policies where the underlying architecture is based on Graph Neural Networks or Transformers.

Format

A seminar with 7 graded ECTS points, run in block format: an initial meeting in November and a block of student presentations in March. Each participant read and understood a piece of research, wrote a summary paper in their own words, gave a presentation, and gave detailed feedback on the paper and presentation of a fellow student.

Topics

Area 1: Analysis of General Policies (supervisor: Timo P. Gros)

Simon Ståhlberg, Blai Bonet, and Hector Geffner. Learning general optimal policies with graph neural networks: Expressive power, transparency, and limits.
Simon Ståhlberg, Blai Bonet, and Hector Geffner. Learning generalized policies without supervision using GNNs.
Rostislav Horčík and Gustav Šír. Expressiveness of graph neural networks in planning domains.
Dillon Z. Chen, Sylvie Thiébaux, and Felipe Trevizan. Learning domain independent heuristics for grounded and lifted planning.

Area 2: Training of General Policies (supervisor: Daniel Höller)

Sam Toyer, Felipe Trevizan, Sylvie Thiébaux, and Lexing Xie. Action schema networks: Generalised policies with deep learning.
Or Rivlin, Tamir Hazan, and Erez Karpas. Generalized planning with deep reinforcement learning.
Simon Ståhlberg, Blai Bonet, and Hector Geffner. Learning general policies with policy gradient methods.

Area 3: General Policies using Transformers (supervisor: Nicola J. Müller)

Nicola J. Müller, Pablo Sánchez, Jörg Hoffmann, Verena Wolf, and Timo P. Gros. Comparing state-of-the-art graph neural networks and transformers for general policy learning.
Nicholas Rossetti, Massimiliano Tummolo, Alfonso Emilio Gerevini, Luca Putelli, Ivan Serina, Mattia Chiari, and Matteo Olivato. Learning general policies for planning through GPT models.
Lucas Lehnert, Sainbayar Sukhbaatar, Paul Mcvay, Michael Rabbat, and Yuandong Tian. Beyond A*: Better planning with transformers via search dynamics bootstrapping.
Lili Chen, Kevin Lu, Aravind Rajeswaran, Kimin Lee, Aditya Grover, Michael Laskin, Pieter Abbeel, Aravind Srinivas, and Igor Mordatch. Decision transformer: Reinforcement learning via sequence modeling.

Organization

All information and materials are available in the CMS.

Bridging the Gap – Language Models and Structured Knowledge in AI

Thu, 19 Sep 2024 00:00:00 +0000

Organizers

Dr. Gerrit Großmann
Cennet Oguz, M.Sc.
Dr. Simon Ostermann
Prof. Dr. Verena Wolf

Organization

For any issues regarding the seminar, please e-mail Gerrit Großmann and have [LanguageModelsSeminar2024] (including the brackets) in the subject line.

This block seminar is open to CoLi and CS students (and related courses).
Depending on the study regulations, you need to register in HISPOS/LSF.
CS students must have registered in HISPOS/LSF by June 26! Otherwise we cannot issue a certificate (“Schein”).
If you are a CS student, please use the seminar assignment system to register.
If you are a CoLi student, please write us an email to register.
Please register only if you are available at the time slot of the seminar.
If you want to take the seminar but were not selected by the assignment system, please apply for the waiting list by emailing us.
The seminar takes place on September 19 (Thursday) and 20 (Friday), in person (10:00 sharp, DFKI, D3.2, room Reuse, next to the main entrance).
The seminar language is English.
The seminar earns you 7 ECTS.

Grading

To pass the seminar, you have to attend all sessions and:

give a presentation (CoLi and CS studenets, worth 3 ECTS);
write reports in which you critically examine the topics in the seminar (only CS students, worth 4 ECTS);
participate in discussions during the seminar;

… with a passing grade.

For CS students, the final grade is calculated from the weighted average. Both CoLi and CS students have the opportunity to earn a bonus for submitting an optional practical project. A good project can improve your final grade by 0.3 points (e.g., from 1.7 to 1.3), an excellent project can boost your grade by 0.7 points (e.g., from 2.3 to 1.7).

Topic Overview

This is a block seminar in September 2024, which will be held in cooperation between the Department of Multilingual Technologies (MLT) and Neuro-Mechanistic Modeling (NMM).

Large language models (LLMs) have swiftly become a cornerstone in AI research, capturing the attention of the public as the most accessible gateway to artificial intelligence. Despite their groundbreaking impact, LLMs are not without their imperfections. Notably, the occurrence of hallucinations and limited reasoning capabilities, particularly in specialized domains, remain significant challenges.

This seminar begins by investigating the theoretical foundations of language representations, tracing the evolution of transformers and their progression towards the cutting-edge LLMs we see today. Building on this foundation, the seminar will then explore promising future directions. Special emphasis will be placed on the integration of LLMs with neuro-symbolic reasoning and the enrichment of these models through knowledge graphs and other forms of structured data.

Requirements

We expect no prior knowledge in language modeling. The seminar is open to CS (including related majors) and CoLi students.

Course Structure

Presentation

Identify the key ideas and concepts and give a self-consistent presentation explaining concepts to your fellow students.

The presentation should be about 20 minutes long.

Please submit your your final slides as .pdf until September 18 (23:59, Berlin time) and have [LanguageModelsSeminar2024] Slides Submission in the subject line.

Here are some suggestions for a good presentation (we will use this as a basis for grading the presentations):

Do not try to condense all the content into 20 minutes. Find a suitable level of abstraction and determine what you think is interesting for your fellow students.
Explore and include supplementary material where it seems useful (literature, youtube, GitHub, medium articles, OpenReview, etc.). The papers provided by us can be seen as a starting point.
The goal is to tell a (self-consistent and entertaining) story - not to convince us that you understand the paper.
Put time and effort into creating visualizations and preparing (running) examples (Please do not rely solely on paper screenshots).
Prioritize concreteness, simplicity, and clarity.
Focus on intuition, high-level understanding, and contextualization, not on technical details.
Don’t overcrowd your slides. Try to avoid full sentences and be cautious with bullet points.
Use equations only when necessary; use color-coded equations to improve their readability (example).
Be critical of the authors’ claims, don’t fall for overselling.
Use slide numbers.

Reports

Each CS student is required to write six reports on six different topics of their choosing (this can include your own topic). Each report should be about two to four pages long. The report should contain a short(!) summary of (what you consider to be) the main contribution or most intriguing idea of the papers. Otherwise, you can freely express your own thoughts on the topic. For instance: What did you like/dislike about the papers (both methodically and didactically)? What are connections to other seminar papers? Can you suggest improvements? What do you think is missing?

You can use the Neurips or Springer LNCS Latex template or any other reasonable format (don’t write an abstract). Please use a spell+grammar checker like languagetool or grammarly before submitting. Do not let ChatGPT do the writing for you (it is horrible to read). But you can use LLMs to catch errors or improve formulations.

The time required to write the reports should approximately add up to 4 CP, which translates to 20 hours of work per report (4 CP x 30 hours per CP).

Please email the reports as a single .pdf file (with a reasonable filename) to Gerrit and Cennet and have [LanguageModelsSeminar2024] Report Submission in the subject line. The report deadline is September 28 (23:59, Berlin time).

Practical Project (Bonus)

You can submit a practical project where you apply what you learned in the seminar. The project doesn’t have to be closely related to your presentation topic but should have a loose connection to any seminar topic. If you plan to do this, please discuss it with your supervisor before September 19 and propose your idea.

Practical projects previously submitted in other university courses or already published are not allowed. Plagiarism will result in expulsion. Ensure that the code is reproducible and easy to run, for example, on Colab. Your project also needs to be well documented (e.g., by providing a tutorial notebook), and you need to disclose all the resources you used.

Deadline is October 4 (23:59, Berlin time). We will make all projects publically available. Please email the project as a single .zip file (with a reasonable filename) to Gerrit and Cennet and have [LanguageModelsSeminar2024] Project Submission in the subject line.

Topics

~~Please vote for your preferred topic: here. Deadline is May 25.~~

Topic 1

Topic 2

Name: Self-Attention and Transformer
Student: Mayur Deshmukh
Supervisor: Cennet
Papers:

Topic 3

Topic 4

Topic 5

Name: Properties of LLMs
Student: Vianney Jerry Tchouwa Takou
Supervisor: Cennet
Paper:

Topic 6

Topic 7

Topic 8

Name: Foundations of Knowledge Graph Reasoning
Student: Larisa Ivanova
Supervisor: Gerrit
Paper:
1. Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction
2. Towards Foundation Models for Knowledge Graph Reasoning

Topic 9

Topic 10

Topic 11

Topic 12

Special Topics in Deep Reinforcement Learning

Mon, 01 Apr 2024 00:00:00 +0000

Team

Timo P. Gros, Prof. Dr. Verena Wolf (lecturers)
Joschka Groß, Felix Jahn, Nicola J. Müller (assistants)

Content

Topics

Deep Reinforcement Learning

Rainbow (paper)
A2C & A3C (paper)
PPO (paper)
Model-based RL & MCTS (book, especially chapters 8.1, 8.10 and 8.11; original MCTS paper; paper with more details)
Alpha(Go) Zero (paper, paper 2)

Safe Reinforcement Learning

Conservative Safety Critics for Exploration (paper)
Constrained Policy Optimization (paper)
Safe Reinforcement Learning via Shielding (paper)

Causal Reinforcement Learning

Invariant Causal Prediction for Block MDP (paper)
Causal Reinforcement Learning Using Observational and Interventional Data (paper)
Can Humans Be Out of the Loop? (paper)

Exploration Techniques

Random Network Distillation (paper)
Go-Explore (paper)

Organization

All information and materials are available in the CMS.

Advanced Topics in Diffusion Modeling — From Theory to Implementation

Fri, 27 Oct 2023 00:00:00 +0000

Organizers

Dr. Gerrit Großmann, Prof. Dr. Verena Wolf.

For any issues regarding the seminar, please e-mail Gerrit Großmann and have [DeepDiffusion2023] in the subject line.

In case you are interested in writing a thesis on this topic, please also contact Gerrit Großmann.

Organization

Please use the seminar assignment system to register. Please register only if you are available at the time slot.
The seminar takes place Fridays from 14:15 to 16:00 in room 1.06 (E1.1) (in-person).
The kick-off meeting takes place on Friday, October 27.
The seminar language is English.
The seminar earns you 7 ECTS.
The seminar is eligible for bachelor/master/graduate students of computer science and related courses.

Schedule

Oct 27: Kick-off
Oct 30: Milestone I — Send three topic preferences to Gerrit
Nov 3: No seminar
Nov 10: No seminar
Nov 17: Q&A session topics 1–6
Nov 24: Q&A session topics 7–12
Dec 1: No seminar
Dec 8: Milestone II — Submit slides + Q&A session topics 1–6
Dec 15: Presentation slot 1 (topics 1, 2, 4)
Dec 22: Q&A session topics 7–12 (can be moved)
Jan 5: Presentation slot 2 (topics 5, 6, 7)
Jan 12: Presentation slot 3 (topics 8, 9)
Jan 19: No seminar
Jan 26: Q&A session topics 1–6
Feb 2: Q&A session topics 7–12
Feb 9: Milestone III — Submit & present tutorial notebook
Feb 23: Milestone IV — Submit reviews
Mar 23: Milestone V — Submit final tutorial notebook

Topic Overview

Diffusion models have recently transformed the field of generative deep learning. This seminar will explore this vibrant area of research.

The seminar is targeted at students who already have a background in deep learning (theoretical and practical) and are keen to dive deeper into probabilistic diffusion and related concepts such as stochastic processes and normalizing flows. Experience with diffusion models will be beneficial, but not required. While we do not aim to provide comprehensive coverage of the topic, we have selected certain papers that we find exceptionally engaging, fun, and thought-provoking.

The seminar is equally divided into two segments: a classical part, where students present a concept based on a research paper, and a practical part, where students develop a tutorial notebook inspired by the ideas from that paper.

Requirements

General background knowledge and practical experience in deep learning are strongly recommended. Experience with diffusion models will be helpful but is not necessary.

Grade

To pass the seminar, you have to attend all sessions and:

give a presentation;
write a tutorial notebook;
review three tutorials;
update your tutorial notebook based on the reviews you received;
participate in discussions;

… with a passing grade. The final grade is based on your presentation (40%), tutorial notebook (50%), and reviews (10%). You will fail the whole seminar if you receive a failing grade in any of the three parts. If we are undecided, we will also consider the discussion participation.

Presentation (40%)

Identify the key ideas and concepts and give a self-consistent presentation explaining these concepts to your fellow students.

The presentation should be 15 to 20 minutes long.

Here are some suggestions for a good presentation (we will use this as a basis for grading the presentations):

The goal is to tell a (self-consistent and entertaining) story — not to convince us that you understand the paper.
Put time and effort into creating visualizations and preparing (running) examples.
Prioritize concreteness, simplicity, and clarity.
Focus on intuition, high-level understanding, and contextualization, not on technical details.
Don’t overcrowd your slides. Try to avoid full sentences and be cautious with bullet points.
Explore and include supplementary material where it seems useful (literature, YouTube, GitHub, Medium articles, OpenReview, etc.).
Use equations only when necessary; use color-coded equations to improve their readability (example).
Be critical of the authors’ claims, don’t fall for overselling.
Use slide numbers.
Submit your slides as .pdf.

Tutorial Notebook (50%)

Your task is to create a self-consistent tutorial within a Jupyter Notebook that elucidates the key concepts of the selected paper through code examples. The scope of this project may vary depending on the complexity of the paper. A complete re-implementation may be achievable or it may be more appropriate to concentrate on a small toy problem focusing on a single concept. The Notebook should be well-structured, containing text and figures that clarify the concepts, as well as well-documented code. For guidance and inspiration, you may refer to projects like the Annotated Diffusion Model and the TeachOpenCADD platform, for instance, E(3)-invariant Graph Neural Networks. In addition, you can check out the Stanford CS224W Graph ML Tutorials.

Formalities:

The tutorial notebook should run on Google Colab and utilize PyTorch.
We will invite each student to explain their code in an individual meeting. You should be able to explain the code in detail and justify design choices.
Submit by uploading to MS Teams with naming scheme Topic03_Grossmann_GeometricLatentDiffusionModelsFor3DMoleculeGeneration.ipynb.
You can download external data (weights, training data) from within your notebook.
Plagiarism will get you expelled from the seminar and potentially exmatriculated. However, you can discuss solutions with your fellow students and take inspiration from open-source implementations.
You can also copy&paste small code snippets if they are clearly indicated.
Comment your code extensively and assign proper names to your functions, variables, and classes.
Ensure to set seeds to make your code deterministic, and remember to save the final weights for reproducibility.
Include the URL of the original paper, your name, and any external resources you use.
The notebooks will be made publicly available.
The presentation of your notebook should take about 5 minutes.
The deadline for submissions is 23:59 local time.

Reviews (10%)

We will assign three tutorial notebooks to each student to review. The primary objective of this review is to offer suggestions for enhancements and to identify any potential mistakes or ambiguities, whether they be technical, conceptual, or grammatical in nature. Each review should be approximately one to two pages in length and must be sent to the author and Gerrit via email.

Topics

Topic	Student	Paper
1	Salaheldin Y. A. Mohamed	Structured Denoising Diffusion Models in Discrete State-Spaces
2	–	Efficient and Degree-Guided Graph Generation via Discrete Diffusion Modeling
3	–	Geometric Latent Diffusion Models for 3D Molecule Generation
4	Bartłomiej Pogodziński	Consistency Models
5	Akansh Maurya	TrojDiff: Trojan Attacks on Diffusion Models with Diverse Targets
6	Davronbek Islamov	Training Diffusion Models with Reinforcement Learning
7	Yasin Esfandiari	Graphically Structured Diffusion Models
8	Soumava Paul	Generative Modelling with Inverse Heat Dissipation
9	Monseej Purkayastha	Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise
10	–	Equivariant flow matching
11	–	Generative Modeling with Optimal Transport Maps
12	–	Improving and Generalizing Flow-Based Generative Models with Minibatch Optimal Transport

AI for the Social Good

Fri, 21 Apr 2023 00:00:00 +0000

Organizers

Dr. Gerrit Großmann, Lisa Dargasz, Kevin Baum, Sarah Sterz, Prof. Verena Wolf.

Please use the seminar assignment system to register. You can also email us to register for the waiting list.

For any issues regarding the seminar, please e-mail Gerrit Großmann and have [SocialGoodSeminar] in the subject line.

The kick-off meeting will be on Friday, April 21, at 2 PM (sharp) (this is also the time slot of the seminar during the semester).

Update: The 2023 edition has come to an end. We were privileged to witness the development of three inspiring projects:

Safe Streets: A pedestrian navigation system designed to prioritize safe routes.
Disaster Tweet: A web interface / browser plugin that identifies tweets referencing authentic natural or man-made disasters.
Simple English: A browser plugin that generates summaries of website content in straightforward, easy-to-understand English.

We would like to thank all participating students for their enthusiasm and dedication throughout the seminar.

Topic Overview

Are you an AI enthusiast eager to make a real difference beyond just theoretical knowledge? Do you aspire to use the latest AI technologies to tackle some of humanity’s most pressing issues and make the world a better place? If so, we invite you to join us for “AI for the Social Good”.

Our goal is to equip you with both theoretical and practical knowledge to create AI solutions to real-world problems. You’ll have the opportunity to collaborate with other participants in a hands-on project to make a difference in your community or the world at large.

Organization

The seminar will take place in person.
The seminar language is English.
The seminar earns you 7 ECTS.
The seminar is eligible for bachelor/master/graduate students (of computer science and related courses).
Registration/unregistration in HISPOS/LSF is due by May 12.
We will meet in room 1.06 (E1.1).

How to Pass

In order to pass the seminar, you need to:

Pass all milestones.
Attend all sessions.
Read papers from the compulsory reading list (resp. watch videos).
Actively contribute to the group work.

Schedule

April 21: Kick-off meeting and group assignment.
May 5: Lecture by Sarah Sterz.
May 9 (Tuesday!): Guest lecture Collected for Profit, Repurposed for Social Good: What Advertising Data Reveals About Society by Ingmar Weber, 14:15 – 16:00, in lecture hall 003 (E1.3).
May 12: In Milestone I, we have a pitch meeting where each group presents three project ideas (we will choose one project to proceed with).
No lecture on May 19 and May 26.
June 2: Informal meeting to discuss Milestone II.
June 9: In Milestone II, each group shows the current status of their projects.
June 16, June 30, July 14: Informal meetings to discuss technical details.
July 21: In Milestone III, each group gives a final presentation on their project.
July 28: In Milestone IV, each group submits a final report.

On Fridays without a Milestone, you will either meet with your supervisor, or we will have a guest lecture, or there will be no meeting at all.

Grading

Each student will receive an individual grade. Grading is based on the work of the group (e.g., how suitable are the project ideas) and the work of the student within the group (e.g., indicated by group discussions or GitLab contributions). More specific requirements and evaluation criteria are listed below for each Milestone.

Milestone I — Coming up with project ideas:

Every group is expected to come up with three project ideas. We will pick one of these for the group to proceed. Your task is to pitch these project ideas and convince us they all fulfill the requirements.

For once, each project should be “socially good” in a reasonable sense. That is, it can likely be used to prospectively impact society (or a subset of society) in a positive and meaningful way. Moreover, each project should be related to artificial intelligence, deep learning, or data science. That is, you work with data that you find, create or collect, and you infer knowledge from this data.

Your project idea does not have to be completely novel. You can also extend or improve an existing project, provide an open-source alternative to a commercial product, or adapt an existing project to specific challenges.

In the pitch meeting, you should be able to answer the questions:

Where does the data come from, and what do you want to do with it?
What is the (degree of) novelty?
Do you contribute to an existing project or do something new?
What similar projects can you learn from, and what is the most relevant scientific literature?
What resources do you need from us? (compute resources, robots, licenses, etc.)

The presentation should be around 5 minutes per project.

Milestone II — Finalizing your strategy:

Milestone II should fall into the period where you have finalized your ideas and start implementing your project. Ideally, you can already present a prototype implementation or solution to a toy problem. The presentation allows you to receive some last-minute feedback and critique before completely committing to your plan. You should be able to answer the questions:

What is your concrete plan to finish the project?
What have you identified as the main challenges?
Where do you need further input?
What ethical concerns do you already have?
How do you plan to finish the project in the given time interval?

The presentation should be around 10 minutes.

Milestone III — Present your results:

Milestone III is the final presentation at the end of the semester that rounds up everything: What have you done? What did work, and what did not? Was your project an overall success or failure? What have you learned, and what would you do differently next time? The presentation should be around 20 minutes.

Milestone IV — Reflect on your impact and contextualize your project:

Milestone IV consists of a group submission and an individual submission.

In the 5–10 pages group report, you should

provide a detailed technical description of what you did, and
judge your project from an ethical point of view: Is it actually social good? In which sense of social good? Can it be misused? You can use the compulsory reading list as guidance.

In addition, all participants must turn in an individual report (roughly one page) on their own explaining what work they did in the project and what they learned. You can also provide overall feedback on the seminar.

Workload:

You will receive seven credit points (CP) for the seminar. Each CP corresponds to a workload of 30 hours. To give you a rough idea of our expectations, we suggest the following time management: Spend 1–2 CP (30–60 hours) on the project proposals, 3–5 CP (90–150 hours) on the actual implementation and attendance, and 1 CP (30 hours) on the final report and the presentations.

Resources

Compulsory reading/watch list:

Everybody has to read/watch at least three papers/videos. In each group, each paper/video must be read/watched by at least one person. Do this before Milestone II.

Texts:

Videos:

Further optional reading:

Example datasets:

Recall that you can always collect/label your own datasets.

Example projects and ideas:

Additionally, correlaid, the DSSG repo list, and datakind might provide useful ideas. We also suggest that you draw on your own life experience for inspiration.

Image credit: SpaceX

Deep Generative Diffusion Models

Mon, 09 Jan 2023 00:00:00 +0000

For any issues regarding the seminar, please e-mail Gerrit Großmann and have [DeepDiffusion] in the subject line.

Organizers

Prof. Dr. Verena Wolf, Gerrit Großmann, Dr. Michael Backenköhler.

Organization

The kick-off meeting takes place on Monday, January 9, 14:15 in room 328 (E 1.3). Send three topic preferences (unordered) by Friday 23:59 to Gerrit.
This seminar takes place on March 6 and 7, 2023 in room 1.06 (E1.1) (we start at 9:30!).
The seminar will take place in-person.
The seminar language is English.
The seminar earns you 7 ECTS.
The seminar is eligible for bachelor/master/graduate students of computer science and related courses.

Topic Overview

Diffusion models are a powerful machine learning tool and a thriving area of research. They enable the generation of new samples with extraordinary and unprecedented quality. This seminar explores the theoretical foundations rooted in the theory of stochastic processes and explores their potential in various domains.

Requirements

General background knowledge in deep learning is recommended.

Grade

To pass the seminar, you have to:

attend both sessions;
give a presentation (with a passing grade);
give an ungraded mock presentation;
participate in discussions;
submit reports (with a passing grade);
read the papers and watch the videos on the Compulsory Reading List;
pass the practical project.

The final grade consists of: your presentation (50%), reports (35%), and participation during the seminar discussions (15%). The practical project can be failed or passed. Extraordinary submissions will earn you a bonus on your final grade.

Presentation (50%)

Identify the key ideas and concepts and give a self-consistent presentation explaining these concepts to your fellow students.

The presentation should be about 20 minutes long (no longer than 25 minutes!).

Here are some suggestions for a good presentation (we will use this as a basis for grading the presentations):

The goal is to tell a (self-consistent and entertaining) story — not to convince us that you understand the paper.
Put time and effort into creating visualizations and preparing (running) examples.
Prioritize concreteness, simplicity, and clarity.
Focus on intuition, high-level understanding, and contextualization, not on technical details.
Don’t overcrowd your slides. Try to avoid full sentences and be cautious with bullet points.
Explore and include supplementary material where it seems useful (literature, YouTube, GitHub, Medium articles, OpenReview, etc.).
Use equations only when necessary; use color-coded equations to improve their readability (example).
Be critical of the authors’ claims, don’t fall for overselling.
Use slide numbers.

Send the first draft of your slides as a .pdf file to your supervisor 7 days before the presentation.

You will also be required to give an ungraded mock presentation in your group (see topic list) by March 5.

Reports (35%)

Every participant has to write three short reports (two to four pages each). You can choose freely from the twelve topics (you may also choose your own topic). The report should contain a short summary of (what you consider to be) the main contribution or most intriguing idea of the paper. Otherwise, you can freely express your own thoughts on the topic. For instance: What did you like/dislike about the paper (both methodically and didactically)? What are connections to other seminar papers? Can you suggest improvements? What do you think is missing?

The report deadline is 14 days after the last presentation (March 21, 23:59). You can use the NeurIPS template (activate the preprint flag) or any other reasonable format (don’t write an abstract). Please use a spell+grammar checker like LanguageTool or Grammarly before submitting.

Discussions (15%)

Read each presented paper in advance (of course, we do not expect you to understand them entirely) and actively participate in the discussion.

Practical Project (fail/pass/bonus)

You can work on the project alone or in groups of two. The deadline is 28 days after the last presentation (April 11, 23:59). Specifications are here.

Topics

Compulsory Reading List

A: Diffusion models explained in 4-difficulty levels
B: How AI Image Generators Work (Stable Diffusion / Dall-E) — Computerphile
C: What are Diffusion Models?
D: (optional) How diffusion models work: the math from scratch
E: (optional) The Annotated Diffusion Model
F: (optional) CVPR 2022 Tutorial: Denoising Diffusion-based Generative Modeling: Foundations and Applications

Monday (March 6, 2023)

Student (Group, Supervisor)	Topic	Additional Material
Maryam Meghdadi Esfahani (1, GG)	Generative Modeling by Estimating Gradients of the Data Distribution (2019)	[1,2,3]
Mohammad Sadegh Akhondzadeh (2, MB)	Permutation Invariant Graph Generation via Score-Based Generative Modeling (2020)	[1,2,7,8]
Jorge Augusto Calvimontes Robles (3, GG)	Denoising Diffusion Probabilistic Models (2020)	[4,5,6]
Devikalyan Das (1, MB)	Score-Based Generative Modeling through Stochastic Differential Equations (2021)	[1,2,7,F]
Lisa Dargasz (3, GG)	High-resolution image synthesis with latent diffusion models (2022)	[10,11,12,F]
Simone Antonelli (4, MB)	Equivariant diffusion for molecule generation in 3D (2022)	[17,26,27]

Tuesday (March 7, 2023)

Student (Group, Supervisor)	Topic	Additional Material
~~Advait Ganesh Athreya (2, GG)~~	~~DiGress: Discrete Denoising diffusion for graph generation (2022)~~	[7,13,14]
Simon Graf (2, MB)	Structure-based Drug Design with Equivariant Diffusion Models (2022)	[15,16,17,18]
Tim Valentin Kruse (4, GG)	3D Shape Generation and Completion through Point-Voxel Diffusion (2022)	[19,20]
~~Vikram Singh (4, MB)~~	~~Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem (2022)~~	[17,21,22,23]
Shreyash Arya (3, GG)	Make-A-Video: Text-to-Video Generation without Text-Video Data (2022)	[10,12,24,25]
~~Hamayoon Behmanush (1, MB)~~	~~Continuous diffusion for categorical data (2022)~~	[2,27,28]

The group is only relevant for the mock presentation. Your supervisors are either Gerrit Großmann (GG) or Michael Backenköhler (MB).

Supplementary Material

A non-exhaustive list of supplemental materials that we think might be helpful to you is provided here. We encourage you to seek and consult additional resources.

The header image was created using DALL-E with the prompt a robot painting a picture, abstract oil painting.

Ausgewählte Themen für das Lehramt Informatik

Sat, 01 Oct 2022 00:00:00 +0000

tba

Deep Reinforcement Learning

Sat, 01 Oct 2022 00:00:00 +0000

For details, have a look at the CMS.

Deep Learning for Drug Discovery

Thu, 21 Apr 2022 00:00:00 +0000

Organizers

Prof. Dr. Verena Wolf, Gerrit Großmann, Michael Backenköhler.

Prof. Dr. Olga Kalinina, Ilya Senatorov (ilya.senatorov@helmholtz-hips.de), Roman Joeres (roman.joeres@helmholtz-hips.de).

This is a joint seminar with the Deep Learning for Drug Discovery Seminar by Prof. Olga Kalinina’s group.

Please contact Gerrit Großmann for questions and have [DLDDSeminar] in the subject line.

Registration

Please use the seminar assignment system to register. Bioinformatics students should apply for Prof. Kalinina’s seminar.

Topic

The discovery of novel therapeutic drugs is extremely costly and time-consuming. In silico drug discovery can potentially accelerate this process. However, this method requires exploring the vast chemical space to find molecules of interest. High expectations are placed on deep learning methods to simplify this process and provide navigation in the realm of possible molecules.

This seminar focuses on deep learning methods for learning molecular representations, predicting molecular properties and, ultimately, for finding novel molecules with desired properties. Primarily, we examine recent advances in the field of geometric deep learning and graph neural networks.

Organisation

This seminar takes place Thursdays 10:15 to 11:45.
A kick-off meeting will be held on Thursday, April 21, 2022.
The first session will be held two weeks later (Thursday, May 5, 2022).
The seminar will take place in-person in E2.1 (ZBI) Room 001.
The seminar language is English.
The seminar earns you 7 ECTS.
The seminar is eligible for bachelor and master students of computer science and related courses.
The seminar includes 10 sessions (with 2 presentations each).

Requirements

Prior knowledge of biochemistry is not expected. General background knowledge in deep learning is recommended.

Grade

To pass, you have to attend all sessions (in person), give a presentation, participate in discussions, and submit reports.

The final grade consists of: your presentation (50%), reports (30%), and participation during the seminar discussions (20%).

Presentation (50%)

Identify the key ideas and concepts and give a self-consistent presentation explaining these concepts to your fellow students. The presentation should focus on intuition and high-level understanding and contextualization, not on technical details. We encourage the students to read related work and explore supplementary material and include it in the presentation where it seems useful (e.g., from related literature, YouTube, GitHub, Medium articles, OpenReview, etc.).

Send a first draft of your slides as a .pdf file to your supervisor 7 days before the presentation (deadline is Thursdays 23:59).

The presentation should be 25 minutes long.

Suggestions for a good presentation:

The goal is to tell a self-consistent and entertaining story — not to convince us that you understand the paper.
Put time and effort into creating visualizations and working out (running) examples.
Prioritize concreteness, simplicity, and clarity.
Don’t overcrowd your slides. Avoid full sentences and be cautious with bullet points.
Use equations only when necessary; use color-coded equations to improve their readability.
Be critical of the authors’ claims, don’t fall for overselling.

Reports (30%)

Every participant has to write a short report for each paper in their group (except their own). Each report should be about two to three pages long. The report should contain a short summary of (what you consider to be) the main contribution or most intriguing idea of the paper. Otherwise, you can freely express your own thoughts on the topic. For instance: What did you like/dislike about the paper (both methodically and didactically)? What are connections to other seminar papers? Can you suggest improvements? What do you think is missing?

The report deadline is 14 days after the last presentation (August 4, 23:59).

You can use the NeurIPS or Springer LNCS LaTeX template or any other reasonable format (don’t write an abstract). Please use a spell+grammar checker like LanguageTool or Grammarly before submitting.

Discussions (20%)

Read each presented paper in advance (of course, we do not expect you to understand them fully) and actively participate in the discussion. We expect each participant to attend each session.

Content

Compulsory Reading List

Background Knowledge

Paper List and Time Table

Date	Type	Student (Advisor)	Topic	Group
05.05.2022	B	Roshni Biswas (Roman)	Convolutional Networks on Graphs for Learning Molecular Fingerprints	D
05.05.2022	CS	Advait Ganesh Athreya (GG)	Neural Message Passing for Quantum Chemistry	D
12.05.2022	B	Mohammadmahdi Koochali (Roman & Ilya)	SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient	A
12.05.2022	B	Pegah Einaliyan (Roman & Ilya)	Objective-Reinforced Generative Adversarial Networks (ORGAN) for Sequence Generation Models	A
19.05.2022	CS (MB)	Shreyash Arya	Junction Tree Variational Autoencoder for Molecular Graph Generation	A
19.05.2022	CS (GG)	Ashwath Pravin Shetty	MolGAN: An implicit generative model for small molecular graphs	A
26.05.2022	–	–	No session – Christi Himmelfahrt	–
02.06.2022	B	Nancy Singh (Roman)	Hit and Lead Discovery with Explorative RL and Fragment-based Molecule Generation	B
02.06.2022	CS (GG)	~~August von Liechtenstein~~	~~Learning to Extend Molecular Scaffolds with Structural Motifs~~ (presentation did not take place, write a report anyway)	C
09.06.2022	B	Joel Chavarria Rivera (Ilya)	Model agnostic generation of counterfactual explanations for molecules	B
16.06.2022	–	–	No session – Fronleichnam	–
23.06.2022	CS (MB)	Martin Bálint	Weisfeiler and Lehman Go Cellular: CW Networks	E
23.06.2022	CS (MB)	Kartik Teotia	E(n) Equivariant Graph Neural Networks	E
30.06.2022	B	Gilbert El Khouri (Ilya)	Molecular machine learning with conformer ensembles	C
30.06.2022	B	Aleksandra Kushnareva (Ilya)	GeoMol: Torsional Geometric Generation of Molecular 3D Conformer Ensembles	C
07.07.2022	CS	Ibrahim Siddig Ibrahim Eltayeb (MB)	SchNet: A continuous-filter convolutional neural network for modeling quantum interactions	D
07.07.2022	CS	Irem Begüm Gündüz (GG)	Spherical Message Passing for 3D Graph Networks	C
14.07.2022	B	Tanya Malkani (Roman)	Self-supervised Graph Transformer on Large-Scale Molecular Data	E
14.07.2022	CS	Kinaan Aamir Khan (MB)	Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction Prediction	E
14.07.2022	B	Sneha Thomas (Roman)	Spec2Mol: An end-to-end deep learning framework for translating MS/MS Spectra to de-novo molecules (originally planned on 9.6.)	B
21.07.2022	B	Nazligul Keske (Ilya)	Drug-target affinity prediction using graph neural network and contact maps	D
21.07.2022	CS	Mattes Alexander Warning (GG)	Iterative Refinement Graph Neural Network for Antibody Sequence-Structure Co-design	B

Type indicates whether a paper is suggested for bioinformatics (B) or CS (CS) students.

Image credit: Tara Winstead

Exploring Graph Neural Networks

Mon, 19 Apr 2021 00:00:00 +0000

Organisation: Prof. Dr. Verena Wolf, Gerrit Großmann.

Please contact Gerrit Großmann for questions about the seminar and have [GNNSeminar] in the subject line.

Registration

Please use the seminar assignment system to register.

Topic

Graphs and networks are ubiquitous in nature, and graph-structured data naturally occurs in a variety of disciplines including physics, chemistry, neuroscience, biology, social science, and epidemiology.

This seminar addresses the problem of learning from graph-structured data. Traditional machine learning and deep learning techniques are not well suited to process data containing rich relational information. Graph Neural Networks (GNNs), Graph Machine Learning, and Geometric Deep Learning offer an interesting perspective on this challenge and are currently one of the fastest-growing topics in ML research. GNNs can assist in discovering drugs, fighting cancer, detecting neurological disorders, analyzing complex systems, recommending friendships in online social networks, controlling traffic, as well as studying epidemiological processes.

We examine recent developments in graph ML research, study state-of-the-art techniques, discuss their limitations, and explore various applications.

A neat overview of GNNs is given in the Graph Neural Networks with Petar Veličković talk.

Organisation

This seminar is a block seminar and will be given on September 23 and 24, 2021, starting at 10:00 sharp.
An online kick-off meeting will be held on Monday, April 19, at 2 pm (1 hour max).
The seminar language is English.
The seminar earns you 7 ECTS.
The seminar is eligible for bachelor and master students of computer science.
Send three topic preferences (unordered) to Gerrit by Monday, April 26.
Depending on the study regulations, registration/unregistration in HISPOS/LSF is due by Tuesday May 18.
Mock presentation deadline: Friday, Sep 10.
Short reports deadline: Sunday, Oct 24 (by email).

Requirements

A basic understanding of machine learning and deep learning will be helpful but is not mandatory. If you are new to the field, the 3blue1brown Neural Network Playlist might be a good start. For further consultation of general concepts, see the Deep Learning Book by Goodfellow et al. (available online). Additional material on GNNs can be found in the Graph Representation Learning Book by Hamilton (available online). There are many more good introductory GNN and ML talks on YouTube.

Grade

To pass, you have to attend all sessions, give a presentation, participate in discussions, and submit reports. Furthermore, we expect all participants to read all seminar papers in the course of the semester and write down questions/thoughts/ideas/etc. (of course, we do not expect you to understand them fully). Each participant must also give an ungraded mock presentation to their group (A, B, or C) at least 2 weeks prior to the actual presentation.

The final grade consists of: your presentation (40%), reports (40%), and participation during the seminar discussions (20%).

Presentation

Identify the key ideas and concepts and give a self-consistent presentation explaining these concepts to your fellow students. The presentation should focus on intuition and high-level understanding and contextualization, not on technical details. We encourage you to read related work and explore supplementary material and include it in the presentation where it seems useful (e.g., from related literature, YouTube, GitHub, Medium articles, OpenReview, etc.). The presentation should be 20 to 30 minutes long.

Report

Each student has to submit four short reports (one for each paper of their group, 1000–2000 words). The report should contain a short summary of (what you consider to be) the main contribution or most intriguing idea of the paper. Otherwise, you can freely express your own thoughts on the topic. For instance: What did you like/dislike about the paper (both methodically and didactically)? What are connections to other seminar papers? Can you suggest improvements? What do you think is missing? Can you think of other useful applications?

You can use the Springer LNCS LaTeX template (every other reasonable formatting is fine, too).

Discussions

Read each presented paper in advance (of course, we do not expect you to understand them fully) and actively participate in the discussion. We expect each participant to attend each session.

Topics

Foundations

Date	Timeslot	Student	Topic	Group
23.09.2021	10:00	Safya Alzayat	Semi-Supervised Classification with Graph Convolutional Networks	A
23.09.2021	10:45	Suruthai Noon Goldstein	Inductive Representation Learning on Large Graphs	B
23.09.2021	–	Siwen Chen	Neural Message Passing for Quantum Chemistry	C

Applications

Date	Timeslot	Student	Topic	Group
23.09.2021	11:30	Gilbert El Khoury	BrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis	A
23.09.2021	13:00	Divya Nidadavolu	Fake News Detection on Social Media using Geometric Deep Learning	B
23.09.2021	13:45	Roman Joeres	Modeling polypharmacy side effects with graph convolutional networks	C
24.09.2021	10:00	Maryam Meghdadi Esfahani	Neural Relational Inference for Interacting Systems	A
24.09.2021	10:45	Yan Yan Lau	Finding Patient Zero: Learning Contagion Source with Graph Neural Networks	B
24.09.2021	11:30	Aakash Rajpal	MolGAN: An implicit generative model for small molecular graphs	C

Properties, limitations, and recent advances

Date	Timeslot	Student	Topic	Group
24.09.2021	12:45	Adarsh Jamadandi	How Powerful are Graph Neural Networks?	A
24.09.2021	13:30	Vabuk Pahari	Strategies for Pre-training Graph Neural Networks	B
24.09.2021	14:15	Zhifei Li	Adversarial Attacks on Neural Networks for Graph Data	C

The participants are split into three groups (A, B, C — with four participants each). Each student writes four reports, one for each paper of their group. The mock presentations are also given within these groups.

Deep Reinforcement Learning

Thu, 01 Apr 2021 00:00:00 +0000

Team

Prof. Dr. Verena Wolf
Timo P. Gros, Joschka Groß (assistants)

The course pages of this seminar were on the group’s former CMS, which is no longer online. The seminar ran again in Winter 2022/23.

Monte Carlo Methods

Fri, 06 Nov 2020 00:00:00 +0000

Organisation: Prof. Dr. Verena Wolf, Michael Backenköhler.

Registration

Please use the seminar assignment system to register.

Topic

The core idea of Monte Carlo (MC) methods is to perform computer simulations of a real-world system based on pseudo-random numbers. MC is applied in a large variety of research areas: physical sciences, computer sciences, engineering, statistics, finance, etc. Moreover, ideas that have originally been developed in the context of MC are meanwhile also in use for sampling problems in the area of machine learning. Although being a simple and very direct approach to analyze a system, MC methods come with a lot of challenges, in particular when sampling rare events or when systems have multiple time scales. In this seminar, we take a computer science perspective on Monte Carlo and cover different MC algorithms to tackle these challenges.

Date	Topic
7. 12.	Variance Reduction
14. 12.	Random Process Generation
4. 1.	Probabilistic Programming
11. 1.	Markov Chain Monte Carlo
18. 1.	Stochastic Gradient Descent
25. 1.	Rare Event Simulation and Cross-Entropy Method
1. 2.	Evolutionary Algorithms

Organisation

All seminar meetings take place online (no physical meetings).
A kick-off meeting takes place on November 6 at 11:00.
The seminar takes place Mondays at 9:00 sharp. A list of dates and topics can be found above and in the CMS.
The seminar language is English.
The seminar earns you 7 ECTS.

Grade

The final grade consists of: your presentation (50%), reports (40%), and participation during the seminar discussions (10%).

Requirements

You should have successfully passed MfI3, the Statistics Lab, or a comparable course covering probability theory/statistics.

Exploring Complex Networks Dynamics

Fri, 18 Oct 2019 00:00:00 +0000

Organisation: Prof. Dr. Verena Wolf, Gerrit Großmann, Michael Backenköhler.

Please contact Gerrit Großmann for questions about the seminar.

Registration

Please use the seminar assignment system to register.

Topic

Many real-world phenomena like rumor spreading, contagions in financial markets, epidemic outbreaks, or cognitive processes can be expressed in the terms of complex networks. The seminar explores dynamical processes which are linked to networked structures in intriguing ways. In particular, we aim at understanding the complex interplay between the topology and the emerging dynamics of a network. We analyze brain networks or online social networks and explore various techniques for their analysis and classification ranging from diffusion models and stochastic simulations to deep learning approaches.

Organisation

All sessions take place in room 328 (E 1.3).
A kick-off meeting will take place on Friday, 18 October, 12:30.
The seminar takes place Thursdays 10:15 – 11:45 with two presentations per session (2 × 20–30 minutes + discussion).
Dates: 14.11.19, 28.11.19, 19.12.19, 9.01.20, 23.01.20, 06.02.20.
The seminar language is English.
The seminar earns you 7 ECTS.

Requirements

There are no prerequisites to take the seminar but a basic understanding of linear algebra and probability theory will be useful.

Grade

The final grade consists of: your presentation (40%), reports (40%), and participation during the seminar discussions (20%).

Presentation

Identify the key ideas and concepts and give a self-consistent presentation explaining these concepts to your fellow students. Assume no prior knowledge about the topic. Read related work and include it where it seems useful to you or where it is necessary to understand the topic. Meet with your supervisor and send her your slides at least one week before your presentation (deadline is Wednesdays 23:59). The duration should be 20–30 min.

Report

In each seminar session (including the session of your own presentation), choose one of the two presented papers/topics and write a short report (600–1600 words) about it. Shortly summarize (what you consider to be) the main contribution or most intriguing idea of the paper. Explain what you liked/disliked about the paper (both methodically and presentation-wise). Highlight connections to other seminar papers. We do not expect comprehensive understanding of the paper.

You can use the Springer LNCS LaTeX template (every other reasonable formatting is fine too).

Send the reports as .pdf files to Gerrit Großmann by email within 13 days of the presentation. Deadline is always Wednesdays 23:59.

Each participant gets one session where they can skip the report. With six sessions, we expect each participant to submit five short reports (but at most one report per session).

Discussions

Read each presented paper in advance (of course, we do not expect you to understand them fully) and actively participate in the discussion. We expect each participant to attend each session.

Topics

Session	Date	Student	Topic
1	14.11.19	Lukas Wachter	Cascading Effects of Targeted Attacks on the Power Grid
1	14.11.19	Redion Xhepa	Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting
2	28.11.19	Christina Eimer	Network control principles predict neuron function in the Caenorhabditis elegans connectome
2	28.11.19	Julian Zimmerlin	BrainNetCNN: Convolutional neural networks for brain networks; towards predicting neurodevelopment
3	19.12.19	Timon Ulrich	Structural Deep Brain Network Mining (Report deadline: 08.01.20)
3	19.12.19	Lisa Heidmann	The epidemic spreading model and the direction of information flow in brain networks (Report deadline: 08.01.20)
4	9.01.20	Sabrina Lauer	Reconstructing an Epidemic Over Time
4	9.01.20	Bastian Herra	Spy vs. Spy: Rumor Source Obfuscation
5	23.01.20	Adrian Dapprich	Bots in Nets: Empirical Comparative Analysis of Bot Evidence in Social Networks (also consider DeBot: Twitter Bot Detection via Warped Correlation)
5	23.01.20	—	~~Dynamic Network Model from Partial Observations~~
6	06.02.20	Nicolas Faroß	Spectral Measures of Distortion for Change Detection in Dynamic Graphs
6	06.02.20	Saleh Al Mohamad	Diffusion-Convolutional Neural Networks

Rein4cement Learning

Tue, 01 Oct 2019 00:00:00 +0000

Organisation: Prof. Dr. Verena Wolf, Timo P. Gros.

Topic

Reinforcement Learning is — besides Supervised and Unsupervised Learning — one of the outstanding variants of Machine Learning. It has evolved to an effective tool, being able to create AI systems that are able to beat humans in almost every game, e.g. the famous strategy board game Go [2].

This seminar will try to do both: teach the basics of Reinforcement Learning in theory and also provide the ability to apply it. We will therefore first have a look at the state-of-the-art algorithms of Reinforcement Learning and afterwards apply these to the well-known game Connect 4.

Image taken from [1].

The seminar will include a short talk, a programming project that will be solved in small groups, as well as a write-up about both: the topics presented in the talk and the implementation.

Course page

Find all materials and the registration details here.

References

Ahmad Hammoudeh, A Concise Introduction to Reinforcement Learning
David Silver, Julian Schrittwieser, Karen Simonyan et al., Mastering the game of Go without human knowledge

Data Networks

Tue, 01 Oct 2013 00:00:00 +0000

Seminar | Neuro-Mechanistic Modeling

Formal Methods for AI Alignment

Team

Content

Requirements

Topics

Organization

Special Topics in Deep Reinforcement Learning

Team

Content

Format

Topics

Organization

Spatiotemporal Models and Inference

Team

Content

Format

Topics

Organization

Hands-On Graph Neural Networks

Lecturers

Organization

Final talktorials

Neuro-symbolic General Policy Learning

Supervisors

Content

Format

Topics

Organization

Bridging the Gap – Language Models and Structured Knowledge in AI

Organizers

Organization

Grading

Topic Overview

Requirements

Course Structure

Presentation

Reports

Practical Project (Bonus)

Topics

Recommended Reading/Watch List

LLM Primer

Project Inspirations and Resources

Special Topics in Deep Reinforcement Learning

Team

Content

Topics

Organization

Advanced Topics in Diffusion Modeling — From Theory to Implementation

Organizers

Organization

Schedule

Topic Overview

Requirements

Grade

Presentation (40%)

Tutorial Notebook (50%)

Reviews (10%)

Topics

Recommended Reading List

AI for the Social Good

Organizers

Topic Overview

Organization

How to Pass

Schedule

Grading

Resources

Deep Generative Diffusion Models

Organizers

Organization

Topic Overview

Requirements

Grade

Presentation (50%)

Reports (35%)

Discussions (15%)

Practical Project (fail/pass/bonus)

Topics

Compulsory Reading List