Search Results for author: Piotr Miłoś

Found 30 papers, 18 papers with code

tsGT: Stochastic Time Series Modeling With Transformer

no code implementations8 Mar 2024 Łukasz Kuciński, Witold Drzewakowski, Mateusz Olko, Piotr Kozakowski, Łukasz Maziarka, Marta Emilia Nowakowska, Łukasz Kaiser, Piotr Miłoś

Time series methods are of fundamental importance in virtually any field of science that deals with temporally structured data.

Time Series

Analysing The Impact of Sequence Composition on Language Model Pre-Training

1 code implementation21 Feb 2024 Yu Zhao, Yuanbin Qu, Konrad Staniszewski, Szymon Tworkowski, Wei Liu, Piotr Miłoś, Yuxiang Wu, Pasquale Minervini

In this work, we find that applying causal masking can lead to the inclusion of distracting information from previous documents during pre-training, which negatively impacts the performance of the models on language modelling and downstream tasks.

In-Context Learning Language Modelling +1

Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation Problem

no code implementations5 Feb 2024 Maciej Wołczyk, Bartłomiej Cupiał, Mateusz Ostaszewski, Michał Bortkiewicz, Michał Zając, Razvan Pascanu, Łukasz Kuciński, Piotr Miłoś

Fine-tuning is a widespread technique that allows practitioners to transfer pre-trained capabilities, as recently showcased by the successful applications of foundation models.

Montezuma's Revenge NetHack +2

MoE-Mamba: Efficient Selective State Space Models with Mixture of Experts

1 code implementation8 Jan 2024 Maciej Pióro, Kamil Ciebiera, Krystian Król, Jan Ludziejewski, Michał Krutul, Jakub Krajewski, Szymon Antoniak, Piotr Miłoś, Marek Cygan, Sebastian Jaszczur

State Space Models (SSMs) have become serious contenders in the field of sequential modeling, challenging the dominance of Transformers.

Exploring Continual Learning of Diffusion Models

no code implementations27 Mar 2023 Michał Zając, Kamil Deja, Anna Kuzina, Jakub M. Tomczak, Tomasz Trzciński, Florian Shkurti, Piotr Miłoś

Diffusion models have achieved remarkable success in generating high-quality images thanks to their novel training procedures applied to unprecedented amounts of data.

Benchmarking Continual Learning +1

Magnushammer: A Transformer-Based Approach to Premise Selection

no code implementations8 Mar 2023 Maciej Mikuła, Szymon Tworkowski, Szymon Antoniak, Bartosz Piotrowski, Albert Qiaochu Jiang, Jin Peng Zhou, Christian Szegedy, Łukasz Kuciński, Piotr Miłoś, Yuhuai Wu

By combining \method with a language-model-based automated theorem prover, we further improve the state-of-the-art proof success rate from $57. 0\%$ to $71. 0\%$ on the PISA benchmark using $4$x fewer parameters.

Automated Theorem Proving Language Modelling +1

Trust Your $\nabla$: Gradient-based Intervention Targeting for Causal Discovery

no code implementations NeurIPS 2023 Mateusz Olko, Michał Zając, Aleksandra Nowak, Nino Scherrer, Yashas Annadani, Stefan Bauer, Łukasz Kuciński, Piotr Miłoś

In this work, we propose a novel Gradient-based Intervention Targeting method, abbreviated GIT, that 'trusts' the gradient estimator of a gradient-based causal discovery framework to provide signals for the intervention acquisition function.

Causal Discovery Experimental Design

Disentangling Transfer in Continual Reinforcement Learning

no code implementations28 Sep 2022 Maciej Wołczyk, Michał Zając, Razvan Pascanu, Łukasz Kuciński, Piotr Miłoś

The ability of continual learning systems to transfer knowledge from previously seen tasks in order to maximize performance on new tasks is a significant challenge for the field, limiting the applicability of continual learning solutions to realistic scenarios.

Continual Learning Continuous Control +2

Thor: Wielding Hammers to Integrate Language Models and Automated Theorem Provers

no code implementations22 May 2022 Albert Q. Jiang, Wenda Li, Szymon Tworkowski, Konrad Czechowski, Tomasz Odrzygóźdź, Piotr Miłoś, Yuhuai Wu, Mateja Jamnik

Thor increases a language model's success rate on the PISA dataset from $39\%$ to $57\%$, while solving $8. 2\%$ of problems neither language models nor automated theorem provers are able to solve on their own.

Automated Theorem Proving

Subgoal Search For Complex Reasoning Tasks

1 code implementation NeurIPS 2021 Konrad Czechowski, Tomasz Odrzygóźdź, Marek Zbysiński, Michał Zawalski, Krzysztof Olejnik, Yuhuai Wu, Łukasz Kuciński, Piotr Miłoś

In this paper, we implement kSubS using a transformer-based subgoal module coupled with the classical best-first search framework.

Rubik's Cube

Continual World: A Robotic Benchmark For Continual Reinforcement Learning

1 code implementation NeurIPS 2021 Maciej Wołczyk, Michał Zając, Razvan Pascanu, Łukasz Kuciński, Piotr Miłoś

Continual learning (CL) -- the ability to continuously learn, building on previously acquired knowledge -- is a natural requirement for long-lived autonomous reinforcement learning (RL) agents.

Continual Learning reinforcement-learning +1

Planning and Learning Using Adaptive Entropy Tree Search

1 code implementation12 Feb 2021 Piotr Kozakowski, Mikołaj Pacek, Piotr Miłoś

We present Adaptive Entropy Tree Search (ANTS) - a novel algorithm combining planning and learning in the maximum entropy paradigm.

Emergence of compositional language in communication through noisy channel

no code implementations ICML Workshop LaReL 2020 Łukasz Kuciński, Paweł Kołodziej, Piotr Miłoś

In this paper, we investigate how communication through a noisy channel can lead to the emergence of compositional language.

Uncertainty-sensitive Learning and Planning with Ensembles

1 code implementation19 Dec 2019 Piotr Miłoś, Łukasz Kuciński, Konrad Czechowski, Piotr Kozakowski, Maciek Klimek

The former manifests itself through the use of value function, while the latter is powered by a tree search planner.

Montezuma's Revenge

Uncertainty - sensitive learning and planning with ensembles

1 code implementation25 Sep 2019 Piotr Miłoś, Łukasz Kuciński, Konrad Czechowski, Piotr Kozakowski, Maciej Klimek

Notably, our method performs well in environments with sparse rewards where standard $TD(1)$ backups fail.

Montezuma's Revenge

Expert-augmented actor-critic for ViZDoom and Montezumas Revenge

2 code implementations10 Sep 2018 Michał Garmulewicz, Henryk Michalewski, Piotr Miłoś

We propose an expert-augmented actor-critic algorithm, which we evaluate on two environments with sparse rewards: Montezumas Revenge and a demanding maze from the ViZDoom suite.

Cannot find the paper you are looking for? You can Submit a new open access paper.