Distributed Computing
ETH Zurich

Andreas Plesner

Andreas Plesner

Distributed Computing Group
Computer Engineering and Networks Laboratory (TIK)
Department Electrical Engineering (D-ITET)
Office ETZ G 63
ETH Zurich

Gloriastrasse 35
8092 Zurich
Switzerland

email

Publications

Theses & Labs

Current Theses & Labs

TitleTypeContact/SupervisorAssignedStudent(s)
Differentiable Logic Gate Networks with Classic Computer Vision EncodingSAndreas Plesner,
Till Aczel		FS 26
Event-based data classification with Differentiable Logic Gate NetworksSAndreas Plesner,
Till Aczel		FS 26
Robust Yield Curve Approximation under Sparse Retraining and Non-Stationary Market ConditionsSAndreas PlesnerFS 26
Solving Minesweeper with Differentiable Logic Gate NetworksSDavid Jenny,
Andreas Plesner		FS 26
Towards Gödel Machines: Practical Self-Referential LLM SystemsSDavid Jenny,
Andreas Plesner		FS 26
Detecting Blood Vessel Networks from 3D ScansSAndreas Plesner,
Joël Mathys		HS 25

Past Theses & Labs

TitleTypeContact/SupervisorAssignedStudent(s)
Is the hls4ml Jet Tagging Benchmark Saturated? A Convergent Diagnosis of the Bayes Floor [confidential]BAndreas Plesner,
Till Aczel		FS 26
BitLogic: A Framework for Gradient-Based LUT-Native Neural Networks & Beyond Depth: Scalable Lookuptable Networks [confidential]MTill Aczel,
Andreas Plesner		HS 25
Exploring Normalized Transformers [confidential]MAndreas Plesner,
Till Aczel		HS 25
Detecting Blood Vessel Networks from 3D Scans [confidential]SJoël Mathys,
Andreas Plesner		HS 25
Eigenvector-Masked AutoencodersBTill Aczel,
Andreas Plesner		FS 25
AlphaZero meets ARC: Explorative Capability Acquisition through Goal-Conditioned Adversarial Curriculum Creation [confidential]MAndreas Plesner,
Benjamin Estermann		FS 25
A Survey on the Abstraction and Reasoning Corpus (2019–2025) [confidential]SAndreas PlesnerFS 25
Context-aware In-Car Recommender System [confidential]MAndreas PlesnerFS 25
Deep Differentiable Logic Gate Networks: Neuron Collapse Through a Neural Architecture Search Perspective [confidential]SAndreas Plesner,
Till Aczel		FS 25
A Comprehensive Benchmark for Graph-Based Abstract Reasoning [confidential]MSaku Peltonen,
Andreas Plesner		FS 25
Recurrent Deep Differentiable Logic Gate Networks [confidential]STill Aczel,
Andreas Plesner		FS 25
Effectiveness of Multi-Scale Aggregation for Adversarial RobustnessBAndreas Plesner,
Till Aczel		FS 25
Evaluating AI-Generated Image Detection Across Resolution and Complexity [confidential]BTill Aczel,
Andreas Plesner		FS 25
Adversarial Attacks on Machine LearningSAndreas PlesnerFS 25
Faster DiffLogic [confidential]SAndreas Plesner,
Till Aczel		FS 25
The Impact of Training Data on Adversarial Examples [confidential]BAndreas Plesner,
Till Aczel		FS 25
Light Differentiable Logic Gate Networks [confidential]STill Aczel,
Andreas Plesner		FS 25
Scalability and Expressiveness of the Group-Sum Layer in Differentiable Logic Gate Networks [confidential]SAndreas Plesner,
Till Aczel		FS 25
Failing at Basic Visual Perception but Solving Complex Visual Tests? Evaluating Fundamental Visual Perception Skills of VLMs on Synthetic Datasets [confidential]BAndreas Plesner,
Frédéric Berdoz		FS 25
Benchmarking LLM Sampling Strategies [confidential]BFrédéric Berdoz,
Andreas Plesner		FS 25
Fine-tuning Data Extraction from Large Language Models [confidential]MAndreas Plesner,
Till Aczel		FS 25
rStar-ARC: Process-Preference–Guided MCTS for Stepwise Program Synthesis [confidential]BAndreas Plesner,
Benjamin Estermann		FS 25
Synthetic Data Augmentation in Medical Image Classification: How Architecture Determines UtilityBAndreas Plesner,
Till Aczel		FS 25
Rethinking Adversarial Examples [confidential]MAndreas PlesnerHS 24
Yield Curve Approximation and Prediction with Recurrent Neural Networks [confidential]BDaniela Brauckhoff,
Andreas Plesner		HS 24
Computing with Asynchronous Cellular Automata [confidential]MFlorian Grötschla,
Andreas Plesner		HS 24
Breaking Image CAPTCHAs and designing a new CAPTCHA [confidential]MAndreas PlesnerHS 24
FastFeedForward Networks in Vision Transformers [confidential]SAndreas PlesnerHS 24
Human-aligned Compression for Robust Models [confidential]SAndreas Plesner,
Till Aczel		HS 24
ACORD: An Expert-Annotated Dataset for Contract Clause Retrieval [confidential]MAndreas PlesnerHS 24
Optimizing Inventory Management with Reinforcement LearningMAndreas Plesner,
Benjamin Estermann		HS 24
Stability of Adversarial Examples [confidential]BAndreas PlesnerHS 24
Breaking Google’s reCAPTCHAv3 [confidential]SAndreas PlesnerFS 24
Breaking Google’s reCAPTCHAv3 - Getting good cookies [confidential]SAndreas PlesnerFS 24
Breaking Google’s reCAPTCHAv3 - How are humans classified [confidential]SAndreas PlesnerFS 24
Improving the Data Quality of ConfSearch using LLMsSAndreas PlesnerFS 24
Data Driven Anomaly Detection for Rails using In-service Railway VehiclesBAndreas PlesnerFS 24
Data Driven Anomaly Detection of Springs in In-service Railway Vehicles [confidential]BAndreas PlesnerFS 24
Sybil Detection using Graph Neural Networks [confidential]MAndreas PlesnerFS 24
Automated Vascular Graph Extraction [confidential]MJoël Mathys,
Andreas Plesner		FS 24
Jigsaw Puzzle Solver using Machine Learning [confidential]SAndreas Plesner,
Till Aczel		FS 24
Reconstructing ODEs for Railway Vehicles Using Machine Learning [confidential]BAndreas PlesnerFS 24
Using Physics-Informed Neural Networks to Predict Railway IrregularitiesMAndreas PlesnerFS 24
Best Pokémon TeamBAndreas PlesnerHS 23
I’m a Robot. Solving CAPTCHAs with Advanced Machine Learning Techniques [confidential]SAndreas PlesnerHS 23
Automated Positioning for Smart Lighting [confidential]MJudith Beestermöller,
Andreas Plesner		HS 23
Generative Models for Financial Data: Applications to Initial Margin and its Backtesting [confidential]MAndreas PlesnerFS 23