BEGIN:VCALENDAR VERSION:2.0 PRODID:-//128.220.36.25//NONSGML kigkonsult.se iCalcreator 2.26.9// CALSCALE:GREGORIAN METHOD:PUBLISH X-FROM-URL:https://www.clsp.jhu.edu X-WR-TIMEZONE:America/New_York BEGIN:VTIMEZONE TZID:America/New_York X-LIC-LOCATION:America/New_York BEGIN:STANDARD DTSTART:20231105T020000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RDATE:20241103T020000 TZNAME:EST END:STANDARD BEGIN:DAYLIGHT DTSTART:20240310T020000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RDATE:20250309T020000 TZNAME:EDT END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT UID:ai1ec-20117@www.clsp.jhu.edu DTSTAMP:20240329T063534Z CATEGORIES;LANGUAGE=en-US:Seminars CONTACT: DESCRIPTION:Abstract\nNeural sequence generation systems oftentimes generat e sequences by searching for the most likely sequence under the learnt pro bability distribution. This assumes that the most likely sequence\, i.e. t he mode\, under such a model must also be the best sequence it has to offe r (often in a given context\, e.g. conditioned on a source sentence in tra nslation). Recent findings in neural machine translation (NMT) show that t he true most likely sequence oftentimes is empty under many state-of-the-a rt NMT models. This follows a large list of other pathologies and biases o bserved in NMT and other sequence generation models: a length bias\, large r beams degrading performance\, exposure bias\, and many more. Many of the se works blame the probabilistic formulation of NMT or maximum likelihood estimation. We provide a different view on this: it is mode-seeking search \, e.g. beam search\, that introduces many of these pathologies and biases \, and such a decision rule is not suitable for the type of distributions learnt by NMT systems. We show that NMT models spread probability mass ove r many translations\, and that the most likely translation oftentimes is a rare event. We further show that translation distributions do capture imp ortant aspects of translation well in expectation. Therefore\, we advocate for decision rules that take into account the entire probability distribu tion and not just its mode. We provide one example of such a decision rule \, and show that this is a fruitful research direction.\nBiography\nI am a n assistant professor (UD) in natural language processing at the Institute for Logic\, Language and Computation where I lead the Probabilistic Langu age Learning group.\nMy work concerns the design of models and algorithms that learn to represent\, understand\, and generate language data. Example s of specific problems I am interested in include language modelling\, mac hine translation\, syntactic parsing\, textual entailment\, text classific ation\, and question answering.\nI also develop techniques to approach gen eral machine learning problems such as probabilistic inference\, gradient and density estimation.\nMy interests sit at the intersection of disciplin es such as statistics\, machine learning\, approximate inference\, global optimization\, formal languages\, and computational linguistics.\n \n DTSTART;TZID=America/New_York:20210419T120000 DTEND;TZID=America/New_York:20210419T131500 LOCATION:via Zoom SEQUENCE:0 SUMMARY:Wilker Aziz (University of Amsterdam) “The Inadequacy of the Mode in Neural Machine Translation” URL:https://www.clsp.jhu.edu/events/wilker-aziz-university-of-amsterdam/ X-COST-TYPE:free X-ALT-DESC;FMTTYPE=text/html:\\n\\n
\\nAbstr act
\nNeural sequence generation systems oftentimes generat e sequences by searching for the most likely sequence under the learnt pro bability distribution. This assumes that the most likely sequence\, i.e. t he mode\, under such a model must also be the best sequence it has to offe r (often in a given context\, e.g. conditioned on a source sentence in tra nslation). Recent findings in neural machine translation (NMT) show that t he true most likely sequence oftentimes is empty under many state-of-the-a rt NMT models. This follows a large list of other pathologies and biases o bserved in NMT and other sequence generation models: a length bias\, large r beams degrading performance\, exposure bias\, and many more. Many of the se works blame the probabilistic formulation of NMT or maximum likelihood estimation. We provide a different view on this: it is mode-seeking search \, e.g. beam search\, that introduces many of these pathologies and biases \, and such a decision rule is not suitable for the type of distributions learnt by NMT systems. We show that NMT models spread probability mass ove r many translations\, and that the most likely translation oftentimes is a rare event. We further show that translation distributions do capture imp ortant aspects of translation well in expectation. Therefore\, we advocate for decision rules that take into account the entire probability distribu tion and not just its mode. We provide one example of such a decision rule \, and show that this is a fruitful research direction.
\nBi ography
\nI am an assistant professor (UD) in natu ral language processing at the Institute for Logic\, Language and Computation where I lead the Probabilistic Language Learning group.
\nMy work concerns the design of models and algorithms that learn to represe nt\, understand\, and generate language data. Examples of specific problem s I am interested in include language modelling\, machine translation\, sy ntactic parsing\, textual entailment\, text classification\, and question answering.
\nI also develop techniques to approach general machine l earning problems such as probabilistic inference\, gradient and density es timation.
\nMy interests sit at the intersection of disciplines such as statistics\, machine learning\, approximate inference\, global optimiz ation\, formal languages\, and computational linguistics.
\n\n< p> \n X-TAGS;LANGUAGE=en-US:2021\,April\,Aziz END:VEVENT BEGIN:VEVENT UID:ai1ec-24465@www.clsp.jhu.edu DTSTAMP:20240329T063534Z CATEGORIES;LANGUAGE=en-US:Seminars CONTACT: DESCRIPTION:Abstract\nLarge Language Models (LLMs) have demonstrated remark able capabilities across various domains. However\, it is still very chall enging to build highly-reliable applications with LLMs that support specia lized use cases. LLMs trained on web data often excel at capturing general language patterns\, but they could struggle to support specialized domain s and personalized user needs. Moreover\, LLMs can produce errors that are deceptively plausible\, making them potentially dangerous for high-trust scenarios. In this talk\, I will discuss some of our recent efforts in add ressing these challenges with data-efficient tuning methods and a novel fa ctuality evaluation framework. Specifically\, my talk will focus on buildi ng multilingual applications\, one crucial use case often characterized by limited tuning and evaluation data.\nBio\nXinyi(Cindy) Wang is a research scientist at Google DeepMind working on Large Language Models(LLM) and it s application to generative question-answering. She has worked on multilin gual instruction-tuning for Gemini and multilingual generative models used in Google search. Before Google DeepMind\, Cindy Wang obtained her PhD de gree in Language Technologies at Carnegie Mellon University. During her Ph D\, she mainly worked on developing data-efficient natural language proces sing~(NLP) systems. She has made several contributions in data selection\, data representation\, and model adaptation for multilingual NLP. DTSTART;TZID=America/New_York:20240308T120000 DTEND;TZID=America/New_York:20240308T131500 LOCATION:Hackerman Hall B17 @ 3400 N. Charles Street\, Baltimore\, MD 21218 SEQUENCE:0 SUMMARY:Cindy Wang (Google DeepMind) “Building Data-Efficient and Reliable Applications with Large Language Models” URL:https://www.clsp.jhu.edu/events/cindy-wang-google-deepmind-building-dat a-efficient-and-reliable-applications-with-large-language-models/ X-COST-TYPE:free X-ALT-DESC;FMTTYPE=text/html:\\n\\n\\n
Abstr act
\nLarge Language Models (LLMs) have demonstrated remark able capabilities across various domains. However\, it is still very chall enging to build highly-reliable applications with LLMs that support specia lized use cases. LLMs trained on web data often excel at capturing general language patterns\, but they could struggle to support specialized domain s and personalized user needs. Moreover\, LLMs can produce errors that are deceptively plausible\, making them potentially dangerous for high-trust scenarios. In this talk\, I will discuss some of our recent efforts in add ressing these challenges with data-efficient tuning methods and a novel fa ctuality evaluation framework. Specifically\, my talk will focus on buildi ng multilingual applications\, one crucial use case often characterized by limited tuning and evaluation data.
\nBio
\nXinyi(Cindy) Wang is a research scientist at Google DeepMind working on La rge Language Models(LLM) and its application to generative question-answer ing. She has worked on multilingual instruction-tuning for Gemini and mult ilingual generative models used in Google search. Before Google DeepMind\, Cindy Wang obtained her PhD degree in Language Technologies at Carnegie M ellon University. During her PhD\, she mainly worked on developing data-ef ficient natural language processing~(NLP) systems. She has made several co ntributions in data selection\, data representation\, and model adaptation for multilingual NLP.
\n X-TAGS;LANGUAGE=en-US:2024\,March\,Wang END:VEVENT END:VCALENDAR