mim-uw/MachineLearningSeminar

Seminarium Magisterskie Machine Learning

557 commits

updated Mar 19, 2026

See the code

README

Machine Learning

Seminarium magisterskie MIMUW

Plan spotkań 2025/2026

2025-10-02 spotkanie organizacyjne
2025-10-09
2025-10-16 Ewa Szczurek - t.b.d.
2025-10-23
2025-10-30 Jan Urbanek Subliminal Learning: Language models transmit behavioral traits via hidden signals in data https://arxiv.org/pdf/2507.14805, Łukasz Smoliński Training Large Language Models to Reason in a Continuous Latent Space
2025-11-06 Marek Lisowski: Scaling Scaling Laws with Board Games, Cezary Botta: Learn To be Efficient: Build Structured Sparsity in Large Language Models
2025-11-20 Jakub Kołaczyński: Spurious Rewards: Rethinking Training Signals in RLVR, Ignacy Stępniewski: T2VUnlearning: A Concept Erasing Method for Text-to-Video Diffusion Models
2025-11-27 Mikołaj Garstecki: DeepSeek-OCR: Contexts Optical Compression
2025-12-4 Mateusz Winiarek: The Impact of Depth on Compositional Generalization in Transformer Language Models
2025-12-11 Stanisław Kiedrzyński: GraphCast: Learning skillful medium-range global weather forecasting, Ignacy Kozakiewicz: SnapFusion: Text-to-Image Diffusion Model on Mobile
2025-12-18 Mateusz Boryczka: What Characterizes Effective Reasoning? Revisiting Length, Review, and Structure of CoT, Michał Nowicki Compute Better Spent: Replacing Dense Layers with Structured Matrices
2026-01-08 Marek Lisowski: AnySplat: Feed-forward 3D Gaussian Splatting from Unconstrained Views
2026-01-15 Jan Urbanek: Activation-Informed Merging of Large Language Models, Mikołaj Garstecki: Progressive Data Dropout: An Embarrassingly Simple Approach to Faster Training
2026-01-22 Cezary Botta: Muon: An optimizer for hidden layers in neural networks, Michał Nowicki: Evaluating multiple models using labeled and unlabeled data
2026-02-26 Ignacy Stępniewski: Memory Mosaics at scale, Jakub Kołaczyński Robo-Dopamine: General Process Reward Modeling for High-Precision Robotic Manipulation
2026-03-05 Łukasz Smoliński: Understanding and Mitigating Numerical Sources of Nondeterminism in LLM Inference
2026-03-12 Ignacy Kozakiewicz Back to Basics: Let Denoising Generative Models Denoise
2026-03-19 Mateusz Boryczka: 1000 Layer Networks for Self-Supervised RL: Scaling Depth Can Enable New Goal-Reaching Capabilities, Mateusz Winiarek (Gated Attention for Large Language Models: Non-linearity, Sparsity, and Attention-Sink-Free)[https://arxiv.org/pdf/2505.06708]
2026-03-26 Stanisław Kiedrzyński: Superposition Yields Robust Neural Scaling

Plan spotkań 2024/2025

2024-10-03 spotkanie organizacyjne
2024-10-10 spotkanie organizacyjne
2024-10-17 Kamil Ciebiera - Mamba 2 , Ewa Dobrowolska - Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation Problem
2024-10-24 Stanisław Świątkowski - On the Measure of Intelligence, Paweł Narkiewicz - LiPO: Listwise Preference Optimization through Learning-to-Rank
2024-10-31 Piotr Kamiński - Physics of Language Models
2024-11-07 Przełożone ze względu na MLinPL
2024-11-14 Paweł Wojciechowski - Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE), Maciej Raczuk - Multi-turn Reinforcement Learning from Preference Human Feedback
2024-11-21 Jakub Kołaczyński - Diffusion Models Are Real-Time Game Engines
2024-11-28 Julian Kwiatkowski - Let's Verify Step by Step
2024-12-05 Rafał Powalski (Ingenix.ai) - Drug Design Topics
2024-12-12 -
2024-12-19 Bartłomiej Sadlej, Przemysław Pietrzak Simulating 500 million years of evolution with a language model
2024-01-16-
2024-01-23- Mikołaj Garstecki Tree of Thoughts: Deliberate Problem Solving with Large Language Models
2025-03-06 Kamil Ciebiera, Przemysław Pietrzak: Deepseek
2025-03-13 Paweł Siwak LIMO: Less is More for Reasoning
2025-03-27 Jakub Kołaczyński Diffusion Models already have a Semantic Latent Space
2025-04-03 Bartłomiej Sadlej Rethinking Visual Counterfactual Explanations Through Region Constraint
2025-05-15 Julian Kwiatkowski Learn Beyond The Answer: Training Language Models with Reflection for Mathematical Reasoning
2025-05-29 Łukasz Smoliński Mutual Reasoning Makes Smaller LLMs Stronger Problem-Solvers, Maksym Matviienko Self-Reflection in LLM Agents: Effects on Problem-Solving Performance
2025-06-05 Ignacy Stępniewski Large Language Diffusion Models
2025-06-12 Marek Lisowski Open-Endedness is Essential for Artificial Superhuman Intelligence, Cezary Botta SAM 2: Segment Anything in Images and Videos, Mateusz Boruń Tensor Programs V: Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer

Plan spotkań 2023/2024

2023-10-05 spotkanie organizacyjne
2023-10-12 przykłady projektów magisterskich
2023-10-19 Maksymilian Grochowski VOYAGER: An Open-Ended Embodied Agent with Large Language Models
2023-10-26 -
2023-11-09 Władek Pałucki Reward Design with Language Models, Maciej Sygnowski ProAgent: Building Proactive Cooperative AI with Large Language Models
2023-11-16 Witold Drzewakowski (Eureka), Mateusz Boruń (Universal Approximation Theorem for MLP)
2023-11-23 Franciszek Szarwacki (Can Large Language Models Reason about Program Invariants? Paper, Slides)
2023-11-30 Ewa Dobrowolska Code as Policies: Language Model Programs for Embodied Control, Kamil Jonak (LLM as General Pattern Machines)
2023-12-07 Kamil Ciebiera LLMs and the Abstraction and Reasoning Corpus, Krystian Król Low-Resource Languages Jailbreak GPT-4
2023-12-14 Bartłomiej Sadlej Autogen
2023-12-21
2024-01-11 Przemysław Pietrzak ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate slides
2024-01-18 Maciej Białobrzeski The Reversal Curse:LLMs trained on “A is B” fail to learn “B is A”
2024-01-25 Mikael Jaworowski (zastosowania Random Matrix Theory w przewidywaniu siły wiązania się potencjalnych leków do wybranych receptorów), Maciej Raczuk Discovering faster matrix multiplication algorithms with reinforcement learning(slides)
2024-02-29 spotkanie organizacyjne - semestr letni
2024-03-07 Kamil Ciebiera + Krystian Król: MoE-Mamba + background (Mamba/Switch) Slajdy
2024-03-14 Bartłomiej Sadlej i Stanisław Świątkowski - Consistency Models https://arxiv.org/abs/2303.01469
2024-03-21
2024-03-26 Krzysztof Tomala Solving olympiad geometry without human demonstrations
2024-04-04 Ewa Dobrowolska Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking
2024-04-11 Przemysław Pietrzak i Paweł Wojciechowski AttnLRP: Attention-Aware Layer-wise Relevance Propagation for Transformers Slides
2024-04-18
2024-04-25 Franciszek Szarwacki Fast and Robust Early-Exiting Framework for Autoregressive Language Models with Synchronized Parallel Decoding (Slides)
2024-05-09 Michał Sala Musika! Fast Infinite Waveform Music Generation
2024-05-16 Radosław Jurczak Evaluating the Moral Beliefs Encoded in LLMs slides, Piotr Kamiński GPT-4 IS TOO SMART TO BE SAFE: STEALTHY CHAT WITH LLMS VIA CIPHER
2024-05-23 Witold Drzewakowski NLHF, Mateusz Szymański VampNet: Music Generation via Masked Acoustic Token Modeling
2024-06-06 Władek Pałucki Contrastive Reinforcement Learning, Maks Grochowski Learning Curricula in Open-Ended Worlds
2024-06-13 Szymon Sadkowski Efficient Streaming Language Models with Attention Sinks, Maciej Raczuk Language models can explain neurons in language models

Plan spotkań 2022/2023

2022-10-20 Łukasz Kuciński, Research topics in RL and Deep Learning.
2022-11-03 Maciej Wojtala, Best First Search-based planning algorithms in the Rubik's Cube environment
2022-11-17 Szymon Tworkowski & Szymon Antoniak, Formal Premise Selection With Language Models
2022-12-01 Bartłomiej Krzepkowski, A Global Past-Future Early Exit Method for Accelerating Inference of Pre-trained Language Models; Aleksandra Mysiak, Transformer probing
2022-12-15 Franciszek Szarwacki Make-a-video: Text-to-Video Generation without Text-Video Data; Philip Smolenski-Jensen Bayesian Neural Networks
2023-01-12 Jakub Modrzewski ???; Krzysztof Tomala Structural re-parametrization of neural networks
2023-01-26 Krystian Król, State of knowledge on Automatic Music Transcription; Szymon Sadkowski How Close is ChatGPT to Human Experts
2023-03-09 Spotkanie organizacyjne
2023-03-23 Jakub Modrzewski ConvNeXt - A ConvNet for the 2020s; Propozycje tematów Transformers for ML Seminar
2023-04-13 Igor Witkowski, Learning to Play Atari with the Help of Instruction Manuals
2023-05-04 Mateusz Kaluzny, A Mathematical Framework for Transformer Circuits
2023-05-11 Michał Sala ???
2023-05-18 Michał Tyrolski (Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search)
2023-05-25
2023-06-01 Witold Drzewakowski (TBA)
2023-06-15 Radosław Jurczak (Transformers on protein sequences: how & why); Alek Leśniak (TBA)

Plan spotkań 2021/2022

2021-10-07 Spotkanie organizacyjne. Łukasz Kuciński : Prezentacja tematów z RL-a
2021-10-14 NCBJ NOMATEN Stefanos Papanikolaou : Prezentacja tematów
2021-10-21 TCL - Prezentacja tematów
2021-11-04 Zuzanna Opała "Extreme Classification", Maciej Wojtala "AlphaStar Algorithm and Its Adaptation in Portfolio Optimisation"
2021-11-18 Zuzanna Kwiatkowska, "State of Audio"
2021-12-02 Michał Raszkowski "Hyperparameter optimisation", Szymon Tworkowski & Michał Tyrolski "Hierarchical Transformers Are More Efficient Language Models"
2021-12-16 Maciej Mikuła Enhancing Goal-Conditioned Reinforcement Learning agents via Planning, Szymon Haponiuk "Neural ordinary differential equations"
2022-01-13 Paweł Gora "RL do optymalizacji strategii w modelach nowotworów"
2022-01-27 Cancelled
2022-03-03 Kałużny Mateusz Jakub "???"
2022-03-17 Mysiak Aleksandra "Dependency trees in NLP", Błajda Mateusz "Style GAN"
2022-03-31 Tomala Krzysztof Jakub, Antoniak Szymon "A glimpse into the world of Meta Learning"
2022-04-21 Smolenski-Jensen Philip Alexander
2022-05-05 Król Krystian Kamil "All you need to know about DALL·E 2", Wierzbiński Marcin Józef
2022-05-19 ~~Bochenek Maria Teresa~~, Mysiurska Marta
2022-06-02 Modrzewski Jakub Maksymilian "NAS - Neural Predictor Approach"

Plan spotkań 2020/2021

2020-10-15;
2020-10-22;
2020-11-05; Jacek Cyranka (tematy prac mgr.), Wojciech Jablonski, Neural Architecture Search for Segmentation (na podstawie DARTS, Auto-DeepLab, DCNAS), Zuzanna Kwiatkowska, Deep Learning Based Open Set Acoustic Scene Classification
2020-11-19; Zuzanna Opala(Training a Cooperative Team in Google Football Environment Using Multi-Agent Reinforcement Learning), Maciej Sypetkowski (Latent space editing in GANs & PGGAN, StyleGAN paper scraps)
2020-12-03; Jakub Modrzewski(COVID-19 AI Diagnosis using only Cough Recordings. Presentation, Article), Marta Mysiurska Multi-objective counterfactual explanation
2020-12-17; Michal Raszkowski (Computer Vision for Autonomous Vehicles), Piotr Krzywicki: Deep Learning in robotic manipulation
2021-01-14; ~~Karol Kowalski~~, Marcin Wierzbiński Gravitational-wave detectors with machine learning)
2021-01-28; Philip Smolenski-Jensen (AlphaFold - state-of-the-art protein folding model ), Maciej Mikula Agent57- Outperforming the Atari Human Benchmark
2021-03-04
2021-03-18; Marcin Wierzbiński - LSTM vs GRU; Zuzanna Opała (Robust High-dimensional Memory-augmented Neural Networks)
2021-04-08; Maciej Mikuła - Relational inductive biases, deep learning, and graph networks; Jakub Modrzewski - Stock Market’s Price Movement Prediction With LSTM Neural Networks
2021-04-29; Philip Smolenski-Jensen (CURL - Contrastive Unsupervised Representations for Reinforcement Learning)
2021-05-20; Zuzanna Kwiatkowska (On gradient-based saliency maps in deep learning: presentation, paper);
2021-06-10; Wojciech Jabłoński (TurboDARTS - better, faster, more efficient Neural Architecture Search); Jakub Jasiulewicz - Pretrained Transformers As Universal Computation Engines (paper, paper, presentation), Marta Mysiurska - ?; Piotr Krzywicki – Lessons learned implementing real-world robot learning system

Plan spotkań 2019/2020

2019-10-03
2019-10-10; Umair Rasheed - Volkswagen
2019-10-17; Michał Futrega (Neural Architecture Search) + Jacek Cyranka (prezentacja tematów prac magisterskich)
2019-10-24; Maciej Biernaczyk - AutoGAN: Neural Architecture Search for Generative Adversarial Networks
2019-10-31; Michał Zawalski - Knowledge-free solving the Rubik’s cube
2019-11-07; Krystian Koziatek - Learning the Depths of Moving People by Watching Frozen People + Marek Cygan: prezentacja tematów prac magisterskich
2019-11-14; Mateusz Doliński - NGBoost: Natural Gradient Boosting for Probabilistic Prediction
2019-11-21; Jakub Skorupski - TossingBot: Learning to Throw Arbitrary Objects with Residual Physics | Prezentacja
2019-11-28; Michał Kukuła - Graph-based sparse neural networks for traffic signal optimization + (Paweł Gora: prezentacja tematów prac magisterskich
2019-12-05 Przemek Biecek: prezentacja tematów prac magisterskich & Piotr Miłoś: prezentacja tematów prac magisterskich
2019-12-12; Mateusz Doliński - Wild patterns: Ten years after the rise of adversarial machine learning
2019-12-19 Henryk Michalewski, Mateusz Malinowski - Prezentacja tematów prac magisterskich
2020-01-09 brak seminarium
2020-01-16 Jakub Jasiulewicz - Playing hard exploration games by watching YouTube (paper)
2020-01-23 Kamil Faber - Optymalizacja Hiperparametrów, Tomasz Kurzelewski
2020-02-27 Zuzanna Kwiatkowska - Introduction to Topological Data Analysis
2020-03-05
2020-03-12 Jakub Sieroń - Emergent Tool Use from Multi-Agent Interaction
2020-03-19
2020-03-26 Maciej Sypetkowski -- Presentation, Self-training with Noisy Student improves ImageNet classification
2020-04-02 Jakub Sieroń
2020-04-16 brak seminarium
2020-04-23 brak seminarium
2020-04-30 (10:15-13:15) Michał Kukuła, Jakub Skorupski, Kamil Tokarski - Autoencoders
2020-05-07
2020-05-14
2020-05-21 (10:15-13:15) Zuzanna Kwiatkowska (AudioAI), Mieczysław Krawiarz (ERASER), Krystian Koziatek (Skin cancer detection & ResNeSt), Michał Futrega Pruning - DL for audio, NLP and video
2020-05-28 Piotr Piękos - Mathematical tasks with nlp methods
2020-06-04

Plan spotkań 2018/2019

2018-10-11; Zuzanna Pilat - Learning algorithms with neural gpu
2018-10-18; Bartosz Biskupski (TCL) - prezentacja tematów prac magisterskich
2018-10-25; (seminarium odwołane) Paweł Gora - prezentacja tematów prac magisterskich
2018-11-08; Paweł Gora - prezentacja tematów prac magisterskich
2018-11-15; Sebastian Jaszczur - Distillation and privileged information
2018-11-22; Przemysław Sadownik - Tree-based Pipeline Optimization Tool for Automating Data Science
2018-11-29; Marcin Papierzyński - Opening the black box of Deep Neural Networks via Information
2018-12-13; Piotr Biliński - prezentacja tematów prac magisterskich.
2018-12-20; Michał Zawalski
2018-01-10; Rafał Sadziak - Adversarial attacks and defences
2018-01-17; Michał Łuszczyk - Realistic Evaluation of Deep Semi-Supervised Learning Algorithms.
2018-01-24; Wojciech Mańke - Neural Architecture Search With Reinforcement Learning
2019-01-31;
2019-02-07;
2019-02-28; Paweł Zięcik - Large-Scale Study of Curiosity-Driven Learning
2019-03-07; Michał Kukuła Learning Optimal and Fair Decision Trees for Non-Discriminative Decision-Making
2019-03-14; Jakub Sieroń Thinking Fast and Slow with Deep Learning and Tree Search
2019-03-21; Maciej Biernaczyk
2019-03-28; Adam Dobrakowski
2019-04-04; Sebastian Jaszczur - Concrete Problems in AI Safety
2019-04-11; Jakub Skorupski - Generating Adversarial Malware Examples for Black-Box Attacks Based on GAN
2019-04-18;
2019-04-25; Mikołaj Błaż - Single-Agent Policy Tree Search With Guarantees
2019-05-02;
2019-05-09; Przemysław Sadownik "Layer-based AutoML with evolutionary algorithms"
2019-05-16; Jacek Maksymiuk - “What is Relevant in a Text Document?”: An Interpretable Machine Learning Approach
2019-05-23; Mateusz Doliński - doc2vec
2019-05-30; Piotr Piękos - Hindsight Experience Replay
2019-06-06; Paweł Zięcik - BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
2019-06-13; Krystian Koziatek - Depth map prediction problem overview, Tomasz Kępa - AutoAugment: Learning Augmentation Strategies from Data

Plan spotkań 2017/2018

2017-10-05; Spotkanie organizacyjne
2017-10-12; Propozycje prac magisterskich - NVidia
2017-10-19; Propozycje prac magisterskich - Henryk Michalewski
2017-10-26; Understanding deep learning requires rethinking generalization, https://arxiv.org/abs/1611.03530 , prezentacja
2017-11-02; Distilling the Knowledge in a Neural Network, https://arxiv.org/abs/1503.02531
2017-11-09; "Why Should I Trust You?": Explaining the Predictions of Any Classifier, https://arxiv.org/abs/1602.04938 prezentacja
2017-11-16; Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images, https://arxiv.org/abs/1412.1897
2017-11-23; Visualizing statistical models: Removing the blindfold, http://had.co.nz/stat645/model-vis.pdf
2017-11-30; mlr: Machine Learning in R, http://jmlr.org/papers/v17/15-066.html
2017-12-07
2017-12-14; Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics, https://arxiv.org/pdf/1705.07115.pdf
2017-12-21; Deep Reinforcement Learning: Pong from Pixels, http://karpathy.github.io/2016/05/31/rl/ https://docs.google.com/presentation/d/1lHvatcCaJXit7Uub4pBDOD_s32VZKq5NRZPpxcaDf8M/edit#slide=id.p3
2018-01-11; Scikit-learn & Pandas
2018-01-18; Visualizing and Understanding Convolutional Networks, https://arxiv.org/abs/1311.2901
2018-01-25; A Critical Review of Recurrent Neural Networks for Sequence Learning https://arxiv.org/abs/1506.00019
2018-03-01; A Unified Approach to Interpreting Model Predictions https://arxiv.org/abs/1705.07874
2018-03-08
2018-03-15; Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm https://arxiv.org/abs/1712.01815, Mastering the game of Go without human knowledge https://www.nature.com/articles/nature24270?sf123103138=1
2018-03-22; Playing Atari with Deep Reinforcement Learning https://www.cs.toronto.edu/~vmnih/docs/dqn.pdf oraz Continuous control with deep reinforcement learning https://arxiv.org/abs/1509.02971
2018-03-28; Paweł Góra, zaczynamy 10:45
2018-04-05; brak seminarium
2018-04-12; Dynamic Routing Between Capsules https://arxiv.org/abs/1710.09829
2018-04-19
2018-04-26; Matrix Capsules with EM Routing https://openreview.net/forum?id=HJWLfGWRb
2018-05-10; GA2M - Interpretable Generalized Additive Models (+ applications) - Przemysław Horban
2018-05-17; NeuroSAT - Learning a SAT Solver from Single-Bit Supervision https://arxiv.org/abs/1802.03685
2018-05-24; Anchors: High-Precision Model-Agnostic Explanations https://homes.cs.washington.edu/~marcotcr/aaai18.pdf
2018-06-07; Continuous control with deep reinforcement learning https://arxiv.org/abs/1509.02971

Contributors

(top 30 of 89)

161 commits

87 commits

66 commits

26 commits

Similar repos

CredibleAI/CCAI-research-seminar

Materiały z seminariów prowadzonych w MI^2 DataLabie.

nplan-io/ml_paper_club

A repository of papers that have been presented at nPlan's machine learning paper club

269

gyunggyung/MLLMArxivTalk

[Google Meet] MLLM Arxiv Casual Talk

pmiller10/cambridge-ai

alan-turing-institute/robots-in-disguise

Information and materials for the Turing's "robots-in-disguise" reading group on fundamental AI…

SanDiegoMachineLearning/talks

Presentations other than book club meetings

116

isamu-isozaki/huggingface-reading-group

This repository's goal is to precompile all past presentations of the Huggingface reading group

davidmacmillan/DeepLearningStudyGroup

Papers, code, etc. for the Deep Learning Study Group, which has been meeting weekly since 2016.…