AIConfPaper
← Search

Henryk Michalewski

18 accepted papers

2025

Structured Packing in LLM Training Improves Long Context Utilization

Konrad Staniszewski, Szymon Tworkowski, Sebastian Jaszczur, Yu Zhao, Henryk Michalewski, Łukasz Kuciński +1
AAAI 2025technical

Recent advancements in long-context language modeling have attracted significant attention, yet their practical applications often suffer from suboptimal context utilization. To efficiently address this issue, we introduce the Structured Packing for Long Context, SPLiCe, a method that uses retrieval…

Cited by 8Source ↗PDF ↗code ↗Scholar ↗
2025

Towards Robust Mathematical Reasoning

Thang Luong, Dawsen Hwang, Hoang H Nguyen, Golnaz Ghiasi, Yuri Chervonyi, Insuk Seo +14
EMNLP 2025

Finding the right north-star metrics is highly critical for advancing mathematical reasoning capabilities of foundation models, especially given that existing evaluations are either too easy or only focusing on getting correct short answers. To address these issues, we present IMO-Bench, a suite of

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2024

Beyond Lines and Circles: Unveiling the Geometric Reasoning Gap in Large Language Models

Spyridon Mouselinos, Henryk Michalewski, Mateusz Malinowski
EMNLP 2024finding

Large Language Models (LLMs) demonstrate ever-increasing abilities in mathematical and algorithmic tasks, yet their geometric reasoning skills are underexplored. We investigate LLMs’ abilities in constructive geometric problem-solving, – one of the most fundamental steps in developing human mathemat…

Cited by 6Source ↗PDF ↗code ↗Scholar ↗
2024

Promptbreeder: Self-Referential Self-Improvement via Prompt Evolution

Chrisantha Fernando, Dylan Sunil Banarse, Henryk Michalewski, Simon Osindero, Tim Rocktäschel
ICML 2024poster

Popular prompt strategies like Chain-of-Thought Prompting can dramatically improve the reasoning abilities of Large Language Models (LLMs) in various domains. However, such hand-crafted prompt-strategies are often sub-optimal. In this paper, we present Promptbreeder, a general-purpose self-referenti…

Cited by 200Source ↗PDF ↗Scholar ↗
2023

A Simple, Yet Effective Approach to Finding Biases in Code Generation

Spyridon Mouselinos, Mateusz Malinowski, Henryk Michalewski
ACL 2023findings

Recently, high-performing code generation systems based on large language models have surfaced. They are trained on massive corpora containing much more natural text than actual executable computer code. This work shows that current code generation systems exhibit undesired biases inherited from the…

Cited by 7Source ↗PDF ↗code ↗Scholar ↗
2023

Focused Transformer: Contrastive Training for Context Scaling

Szymon Tworkowski, Konrad Staniszewski, Mikołaj Pacek, Yuhuai Wu, Henryk Michalewski, Piotr Miłoś
NeurIPS 2023poster

Large language models have an exceptional capability to incorporate new information in a contextual manner. However, the full potential of such an approach is often restrained due to a limitation in the effective context length. One solution to this issue is to endow an attention layer with access t…

Cited by 125Source ↗PDF ↗code ↗Scholar ↗
2023

Natural Language to Code Generation in Interactive Data Science Notebooks

Pengcheng Yin, Wen-Ding Li, Kefan Xiao, Abhishek Rao, Yeming Wen, Kensen Shi +6
ACL 2023long

Computational notebooks, such as Jupyter notebooks, are interactive computing environments that are ubiquitous among data scientists to perform data wrangling and analytic tasks. To measure the performance of AI pair programmers that automatically synthesize programs for those tasks given natural la…

Cited by 64Source ↗PDF ↗code ↗Scholar ↗
2023

RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control

Brianna Zitkovich, Tianhe Yu, Sichun Xu, Peng Xu, Ted Xiao, Fei Xia +48
CoRL 2023poster

We study how vision-language models trained on Internet-scale data can be incorporated directly into end-to-end robotic control to boost generalization and enable emergent semantic reasoning. Our goal is to enable a single end-to-end trained model to both learn to map robot observations to actions a…

Cited by 1068Source ↗Scholar ↗
2022

Hierarchical Transformers Are More Efficient Language Models

Piotr Nawrot, Szymon Tworkowski, Michał Tyrolski, Lukasz Kaiser, Yuhuai Wu, Christian Szegedy +1
NAACL 2022findings

Transformer models yield impressive results on many NLP and sequence modeling tasks. Remarkably, Transformers can handle long sequences, which allows them to produce long coherent outputs: entire paragraphs produced by GPT-3 or well-structured images produced by DALL-E. These large language models a…

Cited by 73Source ↗PDF ↗code ↗Scholar ↗
2022

Measuring CLEVRness: Black-box Testing of Visual Reasoning Models

Spyridon Mouselinos, Henryk Michalewski, Mateusz Malinowski
ICLR 2022poster

How can we measure the reasoning capabilities of intelligence systems? Visual question answering provides a convenient framework for testing the model's abilities by interrogating the model through questions about the scene. However, despite scores of various visual QA datasets and architectures, wh…

Cited by 3Source ↗PDF ↗Scholar ↗
2022

Multi-Game Decision Transformers

Kuang-Huei Lee, Ofir Nachum, Sherry Yang, Lisa Lee, C. Daniel Freeman, Sergio Guadarrama +5
NeurIPS 2022accept

A longstanding goal of the field of AI is a method for learning a highly capable, generalist agent from diverse experience. In the subfields of vision and language, this was largely achieved by scaling up transformer-based models and training them on large, diverse datasets. Motivated by this progre…

Cited by 262Source ↗PDF ↗code ↗Scholar ↗
2022

Solving Quantitative Reasoning Problems with Language Models

Aitor Lewkowycz, Anders Johan Andreassen, David Dohan, Ethan Dyer, Henryk Michalewski, Vinay Venkatesh Ramasesh +8
NeurIPS 2022accept

Language models have achieved remarkable performance on a wide range of tasks that require natural language understanding. Nevertheless, state-of-the-art models have generally struggled with tasks that require quantitative reasoning, such as solving mathematics, science, and engineering questions at…

Cited by 815Source ↗PDF ↗Scholar ↗
2021

Measuring and Improving BERT’s Mathematical Abilities by Predicting the Order of Reasoning.

Piotr Piękos, Mateusz Malinowski, Henryk Michalewski
ACL 2021short

Imagine you are in a supermarket. You have two bananas in your basket and want to buy four apples. How many fruits do you have in total? This seemingly straightforward question can be challenging for data-driven language models, even if trained at scale. However, we would expect such generic languag…

Cited by 34Source ↗PDF ↗Scholar ↗
2021

Sparse is Enough in Scaling Transformers

Sebastian Jaszczur, Aakanksha Chowdhery, Afroz Mohiuddin, Lukasz Kaiser, Wojciech Gajewski, Henryk Michalewski +1
NeurIPS 2021poster

Large Transformer models yield impressive results on many tasks, but are expensive to train, or even fine-tune, and so slow at decoding that their use and study becomes out of reach. We address this problem by leveraging sparsity. We study sparse variants for all layers in the Transformer and propos…

Cited by 101Source ↗PDF ↗Scholar ↗
2020

Model Based Reinforcement Learning for Atari

Łukasz Kaiser, Mohammad Babaeizadeh, Piotr Miłos, Błażej Osiński, Roy H Campbell, Konrad Czechowski +8
ICLR 2020spotlight

Model-free reinforcement learning (RL) can be used to learn effective policies for complex tasks, such as Atari games, even from image observations. However, this typically requires very large amounts of interaction -- substantially more, in fact, than a human would need to learn the same games. How…

Cited by 1127Source ↗code ↗Scholar ↗
2020

Simulation-Based Reinforcement Learning for Real-World Autonomous Driving

Błażej Osiński, Adam Jakubowski, Paweł Zięcina, Piotr Miłoś, Christopher Galias, Silviu Homoceanu +1
ICRA 2020poster

We use reinforcement learning in simulation to obtain a driving system controlling a full-size real-world vehicle. The driving policy takes RGB images from a single camera and their semantic segmentation as input. We use mostly synthetic data, with labelled real-world data appearing only in the trai…

Cited by 191Source ↗Scholar ↗
2018

Reinforcement Learning of Theorem Proving

Cezary Kaliszyk, Josef Urban, Henryk Michalewski, Miroslav Olšák
NeurIPS 2018poster

We introduce a theorem proving algorithm that uses practically no domain heuristics for guiding its connection-style proof search. Instead, it runs many Monte-Carlo simulations guided by reinforcement learning from previous proof attempts. We produce several versions of the prover, parameterized by…

Cited by 202Source ↗PDF ↗code ↗Scholar ↗