You can find recordings from past seminars in here
| Guest Speaker | Affiliation | Talk Title | Date |
|---|---|---|---|
| Haikang Deng | UT Austin | Reward-Augmented Decoding: Efficient Controlled Text Generation With a Unidirectional Reward Model | 18 Dec |
| Prasann Singhal | UT Austin | A Long Way to Go: Investigating Length Correlations in RLHF | 13 Nov |
| Mara Finkelstein and Markus Freitag | Google Research | MBR and QE Finetuning: Training-time Distillation of the Best and Most Expensive Decoding Methods | 6 Nov |
| João Sacramento | ETH Zurich | Uncovering mesa-optimization algorithms in Transformers | 30 Oct |
| Elena Voita | Meta AI | Neurons in Large Language Models | 23 Oct |
| Tri Dao | Stanford and Together.AI | FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness | 9 Oct |
| Jason Wei | OpenAI | Scaling, emergence, and reasoning in large language models | 18 Sep |
| Yihong Chen | UCL-NLP Group | Improving Language Plasticity via Pretraining with Active Forgetting | 11 Sep |
| Shayne Longpre | MIT Media Lab | The Flan Collection: Designing Data and Methods for Effective Instruction Tuning | 07 Aug |
| Niklas Muennighoff | HuggingFace | Scaling Data-Constrained Language Models | 31 Jul |
| Javier Ferrando | Universitat Politècnica de Catalunya | Explaining How Transformers Use Context to Build Predictions | 26 June |
| Roberto Dessì | FAIR Paris and UPF | Cross-Domain Image Captioning with Discriminative Finetuning | 5 June |
| Marzena Karpinska | University of Massachusetts Amherst | Large language models effectively leverage document-level context for literary translation, but critical errors persist | 15 May |
| Alireza Mohammadshahi | EPFL | Compression and Distillation of Massive Multilingual NMT Models | 8 May |
| Tom Kocmi and Christian Federmann | Microsoft Research | Large Language Models Are State-of-the-Art Evaluators of Translation Quality | 17 April |
| Lorenz Kuhn | University of Oxford | Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation | 10 April |
| Alexis Ross | Massachusetts Institute of Technology -- MIT CSAIL | TBA | 03 April |
| Wenda Xu | University of California, Santa Barbara | SEScore2: Retrieval Augmented Pretraining for Text Generation | 20 Mar |
| Biao Zhang | University of Edinburgh | Prompting Large Language Model for Machine Translation: A Case Study | 13 Mar |
| Weijia Xu | University of Maryland | Understanding and Detecting Hallucinations in Neural Machine Translation via Model Introspection | 6 Mar |
| Vilém Zouhar | ETH Zurich | Poor Man's Quality Estimation: Predicting Reference-based MT Metrics Without the Reference | 27 Feb |
| Matthew Finlayson | Allen Institute for AI | Comprehensively evaluating LMs as general-purpose math reasoners | 20 Feb |
| Sweta Agrawal | University of Maryland | Generating and Evaluating Machine Translation in Context | 13 Feb |
| Nikita Moghe | University of Edinburgh | Extrinsic Evaluation of Machine Translation Metrics | 6 Feb |
| Catarina Botelho | Instituto Superior Técnico & INESC-ID (HLT) | Speech as Biomarker for Disease Screening | 30 Jan |
| Misha Khalman | Google Brain | Calibrating Sequence likelihood Improves Conditional Language Generation | 23 Jan |
| Mateusz Krubinski | Charles University, Czech Republic | Can Summary Evaluation Benefit from Translation Evaluation? | 16 Jan |
| Javier Ferrando | Universitat Politècnica de Catalunya | Towards opening the black box of neural machine translation: Source and target interpretations of the transformer | 9 Jan |
| Pasquale Minervini | University of Edinburgh | Implicit MLE: Backpropagating Through Discrete Exponential Family Distributions | 19 Dec |
| Giorgios Vernikos | EPFL | Embarrassingly Easy Document-Level MT Metrics | 28 Nov |
| Sara Magliacane | University of Amsterdam | On Causal Representation Learning | 14 Nov |
| Graham Neubig | Carnegie Mellon University | Neuro-Symbolic Language Modeling with Retrieval Automaton | 7 Nov |
| Carolina Scarton | University of Sheffield | One size does not fit all: building NLP models for real-world applications | 24 October |
| Alex Warstadt | ETH Zürich | Artificial neural networks as models of human language learning | 17 October |
| Ivan Vulić | University of Cambridge & PolyAI | Towards Language Technology for a Truly Multilingual World? | 10 October |
| Pierre Colombo | Université Paris Saclay | What are the best systems? New perspectives on NLP Benchmarking | 3 October |
| Ilia Kulikov | New York University | Characterizing and addressing the issue of oversmoothing in neural autoregressive sequence modeling | 26 September |
| Niranjan Balasubramanian | Stony Brook University | What ails multihop QA and how to fix it | 19 September |
| Weijia Shi | University of Washingto | Nearest Neighbor Zero-Shot Inference | 25 July |
| Vikas Raunak and Matt Post | Microsoft | SALTED: A Framework for SAlient Long-Tail Translation Error Detection | 18 July |
| Jungo Kasai | University of Washington | Twist Decoding: Diverse Generators Guide Each Other | 27 June |
| Jonas Pfeiffer | New York University / Technical University of Darmstadt | Lifting the Curse of Multilinguality by Pre-training Modular Transformers | 20 June |
| Edward Hu | Mila | Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer | 30 May |
| Andrey Malinin | Yandex Research | Uncertainty Estimation in Autoregressive Structured Prediction | 16 May |
| Jannis Vamvas | University of Zurich | As Little as Possible, as Much as Necessary: Detecting Over- and Undertranslations with Contrastive Conditioning | 2 May |
| Vânia Mendonça | Instituto Superior Técnico - University of Lisbon | Onception: Active Learning with Expert Advice for Real World Machine Translation | 18 Apr |
| Ana Marasovic | Allen Institute for AI, University of Washington | Self-Explaining for Intuitive Interaction with AI | 11 Apr |
| Anirudh Goyal | Mila, University of Montreal | Coordination Among Neural Modules Through a Shared Global Work space | 4 Apr |
| João Sacramento | ETH Zurich | On biologically-plausible meta-learning | 28 Mar |
| Clara Meister | ETH Zurich | Typical Decoding for Natural Language Generation | 14 Mar |
| Chantal Amrhein | University of Zurich | Identifying Weaknesses in Machine Translation Metrics Through Minimum Bayes Risk Decoding: A Case Study for COMET | 7 Mar |
| Markus Freitag | Google Research | Minimum Bayes Risk Decoding with Neural Metrics of Translation Quality | 28 Feb |
| Tom Kocmi | Microsoft Research | To ship or not to ship? | 21 Feb |
| Caleb Ziems | Georgia Tech | Social NLP for Positive and Civil Interactions | 14 Feb |
| Rémi Leblond | DeepMind | Machine Translation Decoding beyond Beam Search | 07 Feb |
| Elizabeth Salesky | Johns Hopkins University | Robust Open-Vocabulary Translation from Visual Text Representations | 31 Jan |
| Samuel Bowman | New York University | When Combating Hype, Proceed with Caution | 24 Jan |
| Victor Veitch | University of Chicago and Google Research | Counterfactual Invariance to Spurious Correlations | 17 Jan |
| Leonardo Ribeiro | TU Darmstadt | Improving Graph-to-Text Generation with Graph Neural Encoders and Pretrained Language Models | 10 Jan |
| Victor Sanh | HuggingFace | Lots of data, lots of parameters: a different way to get funky task zero-shot generalization | 20 Dec |
| Laurent Besacier | Naver Labs Europe | Self Supervised Representation Learning for Pre-training Speech Systems (and Its Application to Speech Translation) | 13 Dec |
| Albert Gu | Stanford University | Efficiently Modeling Long Sequences with Structured State Spaces | 29 Nov |
| Edoardo Ponti | Mila - Quebec | Inductive Biases for Modelling the World's Languages | 6 Dec |
| Wilker Aziz | University of Amsterdam | Sampling-based minimum Bayes risk decoding for NMT | 22 Nov |
| Junxian He & Chunting Zhou | Carnegie Mellon University | Towards a Unified View of Parameter-Efficient Transfer Learning | 25 Oct |
| Rico Sennrich | University of Zurich | Hallucinations and Minimum Bayes Risk Decoding in Neural Machine Translation | 18 Oct |
| Masha Itkina | Stanford University | Evidential Sparsification of Multimodal Latent Spaces in Conditional Variational Autoencoders | 11 Oct |
| Sabrina Mielke | John Hopkins University | Fair Comparisons and Fundamental Ideas for Open-Vocabulary Generative Language and Translation Models | 27 Sep |
| Clara Meister | ETH Zurich | Language Model Evaluation Beyond Perplexity | 20 Sep |
| Samuel Kiegeland | Heidelberg University | Revisiting the Weaknesses of Reinforcement Learning for Neural Machine Translation | 13 Sep |
| Danish Pruthi | Carnegie Mellon University | Topics on Explainability and Evaluation of Explanations | 06 Sep |