Every Friday 12:00-12:30 UTC+1
We are a research group that gathers every friday to discuss one research paper in the area of AI.
We are interested in any kind of AI applied to any kind of data (text, images, graphs, video, tabular, you name it!).
Paper: Natural Language is All a Graph Needs
Presenter: Filipe Assunção
Date: 15/09/2023
Paper: A Practical Deep Learning-Based Acoustic Side Channel Attack on Keyboards
Presenter: Bartlomiej Matejczyk
Date: 15/09/2023
Paper: Engineering a Formally Verified Automated Bug Finder
Presenter: Arthur Correnson
Date: 16/06/2023
Paper: LIMA: Less Is More for Alignment
Presenter: João Nadkarni
Date: 02/06/2023
Paper: CodeGen2: Lessons for Training LLMs on Programming and Natural Languages
Presenter: João Lages
Date: 26/05/2023
Paper: Target-Driven Compositional Concolic Testing with Function
Summary Refinement for Effective Bug Detection
Presenter: Alexandre Lemos
Date: 19/05/2023
Paper: Deferred concretization in symbolic execution via fuzzing
Presenter: Joana Coutinho
Date: 12/05/2023
Paper: DGM: A deep learning algorithm for solving partial differential equations
Presenter: Bartlomiej Matejczyk
Date: 05/05/2023
Paper: Generative Agents: Interactive Simulacra of Human Behavior
Presenter: André Menezes
Date: 21/04/2023
Paper: CoLT5: Faster Long-Range Transformers with Conditional Computation
Presenter: João Nadkarni
Date: 14/04/2023
Paper: Planning with Large Language Models for Code Generation
Presenter: João Lages
Date: 31/03/2023
Paper: UniLayout: Taming Unified Sequence-to-Sequence Transformers for Graphic Layout Generation
Presenter: Vasco Almeida
Date: 24/03/2023
Paper: Pretraining Language Models with Human Preferences
Presenter: Samuel Arcadinho
Date: 13/03/2023
Paper: Jaint: A Framework for User-Defined Dynamic
Taint-Analyses based on Dynamic Symbolic
Execution of Java Programs
Presenter: Alexandre Lemos
Date: 10/03/2023
Paper: Relational Memory Augmented Language Models
Presenter: Bartlomiej Matejczyk
Date: 03/03/2023
Paper: KLEE: unassisted and automatic generation of high-coverage tests for complex systems programs
Presenter: Joana Coutinho
Date: 24/02/2023
Paper: Hungry Hungry Hippos: Towards Language Modeling with State Space Models
Presenter: André Menezes
Date: 17/02/2023
Abstract: JetStream: Automatic Optimization of Virtual Switch Rulesets
Presenter: Margarida Ferreira
Date: 10/02/2023
Paper: REPLUG: Retrieval-Augmented Black-Box Language Models
Presenter: João Nadkarni
Date: 03/02/2023
Paper: Onception: Active Learning with Expert Advice for Real World Machine Translation
Presenter: Vania Mendonça
Date: 27/01/2023
Paper: Reinforcement Learning from Human Feedback
Presenter: João Lages
Date: 13/01/2023
Paper: A Generalization of ViT/MLP-Mixer to Graphs
Presenter: Samuel Arcadinho
Date: 06/01/2023
Paper: Discovering faster matrix multiplication algorithms with reinforcement learning
Presenter: Bartlomiej Matejczyk
Date: 23/12/2022
Paper: Personalizing and Customizing AI Explanations for Clinicians
Presenter: Francisco Calisto
Date: 16/12/2022
Paper: Concolic Execution for WebAssembly
Presenter: Filipe Marques
Date: 09/12/2022
Paper: Gradient Descent: The Ultimate Optimizer
Presenter: André Menezes
Date: 02/12/2022
Paper: Locating and Editing Factual Associations in GPT
Presenter: João Nadkarni
Date: 25/11/2022
Paper: Arbiter: Bridging the Static and Dynamic Divide in Vulnerability Discovery on Binary Programs
Presenter: Alexandre Lemos
Date: 18/11/2022
Paper: Calibrate Before Use: Improving Few-Shot Performance of Language Models
Presenter: Catarina Belém
Date: 11/11/2022
Paper: GLITCH: Automated Polyglot Security Smell Detection in Infrastructure as Code
Presenter: Nuno Saavedra
Date: 11/04/2022
Paper: Large Language Models are Human-Level Prompt Engineers
Presenter: Jakub Zavrel
Date: 28/10/2022
Paper: Understanding HTML with Large Language Models
Presenter: David Aparício
Date: 21/10/2022
Paper: Explaining NLP Models via Minimal Contrastive Editing (MICE)
Presenter: João Lages
Date: 14/10/2022
Paper: Phenaki: Variable Length Video Generation from Open Domain Textual Descriptions
Presenter: Samuel Arcadinho
Date: 07/10/2022
Paper: egg: Fast and Extensible Equality Saturation
Presenter: Miguel Neves
Date: 30/09/2022
Paper: Beyond neural scaling laws: beating power law scaling via data pruning
Presenter: André Menezes
Date: 23/09/2022
Paper: Diachronic language models
Presenter: Daniel Loureiro
Date: 16/09/2022
Paper: The Web Is Your Oyster - Knowledge-Intensive NLP against a Very Large Web Corpus
Presenter: Sofia Aparicio
Date: 02/08/2022
Paper: Describing Differences between Text Distributions with Natural Language
Presenter: David Aparício
Date: 26/08/2022
Paper: ExpertosLF: Dynamic Late Fusion of CBIR Systems using Online Learning with Relevance Feedback
Presenter: Soraia Meneses Alarcão
Date: 12/08/2022
Paper: CONFETTI: Amplifying Concolic Guidance for Fuzzers
Presenter: Alexandre Lemos
Date: 05/08/2022
Paper: SVTS: Scalable Video-to-Speech Synthesis
Presenter: Rodrigo Mira
Date: 29/07/2022
Paper: Assessing robustness in multimodal transportation systems
Presenter: João Aparício
Date: 22/07/2022
Paper: Emerging Properties in Self-Supervised Vision Transformers
Presenter: Lourenço Vaz Pato
Date: 15/07/2022
Paper: P-Adapters: Robustly Extracting Factual Information from Language Models with Diverse Prompts
Presenter: João Lages
Date: 08/07/2022
Paper: Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding
Presenter: João Nadkarni
Date: 01/07/2022
Paper: CoNT: Contrastive Neural Text Generation
Presenter: Samuel Arcadinho
Date: 24/06/2022
Paper: Simplifying Multilingual News Clustering Through Projection From a Shared Space
Presenter: João Santos
Date: 03/06/2022
Paper: BLITZ: Compositional Bounded Model Checking for Real-World Programs
Presenter: Miguel Neves
Date: 27/05/2022
Paper: SimGRACE: A Simple Framework for Graph Contrastive Learning without Data Augmentation
Presenter: André Menezes
Date: 20/05/2022
Paper: Generating Visual Explanations
Presenter: Isabel Rio-Torto
Date: 13/05/2022
Paper: Chain of Thought Prompting Elicits Reasoning in Large Language Models
Presenter: Mariana Lourenço
Date: 06/05/2022
Paper: Input Invariants
Presenter: Alexandre Lemos
Date: 29/04/2022
Paper: Learning to Scaffold: Optimizing Model Explanations for Teaching
Presenter: Patrick Fernandes & Marcos Treviso
Date: 22/04/2022
Paper: EvoKG: Jointly Modeling Event Time and Network Structure for Reasoning over Temporal Knowledge Graphs
Presenter: David Aparício
Date: 08/04/2022
Paper: LayoutGMN: Neural Graph Matching for Structural Layout Similarity
Presenter: Filipe Assunção
Date: 01/04/2022
Paper: SpeechPainter: Text-conditioned Speech Inpainting
Presenter: Sofia Aparício
Date: 25/03/2022
Paper: Once-for-All: Train One Network and Specialize it for Efficient Deployment
Presenter: João Lages
Date: 18/03/2022
Paper: VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning
Presenter: André Menezes
Date: 25/02/2022
Paper: Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead
Presenter: Mariana Lourenço
Date: 18/02/2022
Paper: Using MaxSAT for Efficient Explanations of Tree Ensembles
Presenter: Alexandre Lemos
Date: 11/02/2022
Paper: Training Language Models to Follow Instructions with Human Feedback
Presenter: David Aparício
Date: 04/02/2022
Paper: Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets
Presenter: João Lages
Date: 28/01/2022
Paper: Improving language models by retrieving from trillions of tokens
Presenter: João Nadkarni
Date: 21/01/2022
Paper: Active learning for imbalanced data under cold start
Presenter: Ricardo Barata
Date: 07/01/2022
Paper: VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text
Presenter: Filipe Assunção
Date: 17/12/2021
Paper: MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontiers
Presenter: Samuel Arcadinho
Date: 10/12/2021
Paper: CORGI: Content-Rich Graph Neural Networks with Attention
Presenter: André Menezes
Date: 26/11/2021
Paper: Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data
Presenter: Mariana Lourenço
Date: 12/11/2021
Paper: LEGO: Latent Execution-Guided Reasoning for Multi-Hop Question Answering on Knowledge Graphs
Presenter: David Aparício
Date: 05/11/2021
Paper: Fuzzing Symbolic Expressions
Presenter: Alexandre Lemos
Date: 29/10/2021
Paper: Time-series Generative Adversarial Networks
Presenter: João Lages
Date: 01/10/2021
Paper: CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generation
Presenter: João Nadkarni
Date: 24/09/2021
Paper: Leveraging Sparse Linear Layers for Debuggable Deep Networks
Presenter: Filipe Assunção
Date: 16/09/2021
Paper: Constraint-Based Multi-Objective Combinatorial Optimization
Presenter: Miguel Neves
Date: 10/09/2021
Paper: GUIGAN: Learning to Generate GUI Designs Using Generative Adversarial Networks
Presenter: David Aparício
Date: 20/08/2021
Paper: Neural Architecture Search without Training
Presenter: Filipe Assunção
Date: 13/08/2021
Paper: Deduplicating Training Data Makes Language Models Better
Presenter: João Lages
Date: 06/08/2021
Paper1: Scout: Rapid Exploration of Interface Layout Alternatives through High-Level Design Constraints
Paper2: GRIDS: Interactive Layout Design with Integer Programming
Presenter: Alexandre Lemos
Date: 30/07/2021
Paper: UniT: Multimodal Multitask Learning with a Unified Transformer
Presenter: Sofia Aparício
Date: 16/07/2021
Paper: Exploring Simple Siamese Representation Learning
Presenter: André Menezes
Date: 09/07/2021
Paper: CodeBERT: A Pre-Trained Model for Programming and Natural Languages
Presenter: João Nadkarni
Date: 02/07/2021
Paper: Synthesizing Object State Transformers for Dynamic Software Updates
Presenter: David Aparício
Date: 18/06/2021
Paper: Knowledge Refactoring for Inductive Program Synthesis
Presenter: Alexandre Lemos
Date: 07/06/2021
Paper: Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning
Presenter: João Lages
Date: 28/05/2021
Paper: Strategies for Pre-training Graph Neural Networks
Presenter: Filipe Assunção
Date: 13/05/2021
Paper: Self-Supervised Learning of Graph Neural Networks: A Unified Review
Presenter: André Menezes
Date: 23/04/2021
Paper: End-to-End Learning of Video Compression Using Spatio-Temporal Autoencoders
Presenter: Jorge Pessoa
Date: 09/04/2021
Paper: On the Fly Synthesis of Edit Suggestions
Presenter: Miguel Neves
Date: 26/03/2021
Paper: Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting
Presenter: João Nadkarni
Date: 12/03/2021
Paper: MTFuzz: fuzzing with a multi-task neural network
Presenter: David Aparício
Date: 12/03/2021
Paper: RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers
Presenter: João Lages
Date: 05/03/2021
Paper: BUSTLE: Bottom-up program-Synthesis Through Learning-guided Exploration
Presenter: José Amaral
Date: 26/02/2021
Paper: Demonstrating the Evolution of GANs through t-SNE
Presenter: Victor Costa
Date: 19/02/2021
Paper: ADnEV: Cross-Domain Schema Matching using Deep Similarity Matrix Adjustment and Evaluation
Presenter: Filipe Assunção
Date: 12/02/2021
Paper: Active Inductive Logic Programming for Code Search
Presenter: David Aparício
Date: 05/02/2021
Paper: Towards federated Learning at scale: system design
Presenter: Filipe Assunção
Date: 29/01/2021
Paper: Peregrine: a pattern-aware graph mining system
Presenter: Miguel Neves
Date: 22/01/2021