1 10 15 https://urss.warwick.ac.uk/files/original/7a50b332024da3bacc61ccca099821cf.pdf 2861b456afc2087411456f62fd6966eb Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Paper An academic paper Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Developing Algorithms to Detect and Classify Floods using Google Earth Engine API and Machine Learning Subject The topic of the resource Computer Science Creator An entity primarily responsible for making the resource Andrew Nicoll Date A point or period of time associated with an event in the lifecycle of the resource 2021 Computer Science Flooding Machine learning https://urss.warwick.ac.uk/files/original/cb9f20e05ba32048997d33acda78d623.pdf 182d800b7fcf9ecc8faf73a068c4d3e0 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Poster An academic poster Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Influencing Norms in Congestion Games Subject The topic of the resource Computer Science Creator An entity primarily responsible for making the resource Avi Semler Date A point or period of time associated with an event in the lifecycle of the resource 2023 Avi Semler Computer Science congestion game graph learning Machine learning reinforcement simulation social transport https://urss.warwick.ac.uk/files/original/ddde9f93828349de18b12b0beb542a6a.pdf 182d800b7fcf9ecc8faf73a068c4d3e0 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Poster An academic poster Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Influencing Norms in Congestion Games Subject The topic of the resource Computer Science Creator An entity primarily responsible for making the resource Avi Semler Date A point or period of time associated with an event in the lifecycle of the resource 2023 Avi Semler Computer Science congestion game Machine learning reinforcement learning simulation social graph transport https://urss.warwick.ac.uk/files/original/adf3da93a6607b235a0fc8d7eca57f7c.pdf 232af1c08288408136a44f42a5055241 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Poster An academic poster Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Searching for Bc+ → π+μ+μ– Decays Subject The topic of the resource Physics Creator An entity primarily responsible for making the resource Callum White Date A point or period of time associated with an event in the lifecycle of the resource 2022 algorithum Machine learning signal decays TMVA XGBoost https://urss.warwick.ac.uk/files/original/ac7d79048eae7e2132401691aa7be07d.pdf d19ca2dc9c8e7d9e23dfc51053ca5e46 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Mathematics Academic Paper An academic paper Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Using Multivariate Convolutional Neural Networks to Predict Future Electricity Grid Frequency Subject The topic of the resource Engineering Creator An entity primarily responsible for making the resource Chris Leigh Date A point or period of time associated with an event in the lifecycle of the resource 2023 AI Chris Leigh electricity Engineering Machine learning Predictions Time series https://urss.warwick.ac.uk/files/original/610ae13496b76d9f127d10c5055b7456.pdf 16d8ac366ff5a3fd66668d341d42a61e Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Poster An academic poster Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource AI's Bridge between Structure and Pattern Subject The topic of the resource Physics Creator An entity primarily responsible for making the resource Joe Webb Date A point or period of time associated with an event in the lifecycle of the resource 2023 A! generated Joe Webb Machine learning Physics https://urss.warwick.ac.uk/files/original/32212e6da1d98a7ef3a88ca202a4ba8b.pdf 8c68705fa054bf92460e971858da9a0d Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Poster An academic poster Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Predicting the solubility of amorphous drugs via machine learning. Subject The topic of the resource Chemistry Creator An entity primarily responsible for making the resource Jonas Miksys Date A point or period of time associated with an event in the lifecycle of the resource 2021 Chemistry Drugs Machine learning https://urss.warwick.ac.uk/files/original/6940729a67f8d51f33b21e31eaf0f790.pdf 011c5223a12a4e40ffaa782bccf66336 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Paper An academic paper Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Development of Machine Learning Algorithms for Determining Mosquito Orientation Subject The topic of the resource Engineering Creator An entity primarily responsible for making the resource Joseph Garratt Date A point or period of time associated with an event in the lifecycle of the resource 2022 Machine learning mosquito resistant https://urss.warwick.ac.uk/files/original/8fc2754163c0e086d7050b0d1bb842f5.pdf 952ec5d66fdc4bf6febc6b24ad8bc9d7 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Paper An academic paper Abstract The abstract of the paper As the adoption of Large Language Models (LLMs) is becoming ubiquitous, understanding their pitfalls is becoming increasingly urgent. LLMs are known to hallucinate and may be subject to cascading hallucinatory generations due to their autoregressive structure and exposure bias. This paper first introduces the topic of hallucination cascade, covering relevant literature. Then an experiment is presented on the propagation of hallucination through long answer text generation, for all combinations of five LLMs - Qwen2 (0.5B), Qwen2 (7B), Llama3.1 (8B), Gemma2 (9B) and Mistral-NeMo (12B) - at five levels of temperature: 0, 0.75, 1, 1.25, 2. In a novel methodology, responses are generated by the five LLMs, then labelled by two more LLMs as hallucinatory or not on both a token and sentence level. The labelling is done with clear instructions and additional context to ensure its quality. Correlational evidence was found on both granularity levels that hallucinatory tokens are positively autocorrelated at lag 1, and that autocorrelation monotonically decreases as lag increases (token level only), supporting the cascading hallucination hypothesis. Finally, some hallucination mitigating techniques are presented, with both research and application in mind. All the code used for the empirical part of this paper (Section 3) can be found at https://github.com/LucMekouar/URSS_2024. Meta Tags Meta tags Large Language Model, LLM, Machine Learning, Natural Language Processing, Hallucination Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Subject The topic of the resource Computer Science Date A point or period of time associated with an event in the lifecycle of the resource 2024 Contributor An entity responsible for making contributions to the resource Prof Bo Chen Title A name given to the resource Investigating Hallucination Cascades in Autoregressive Large Language Models Creator An entity primarily responsible for making the resource Luc Mekouar Hallucination Large Language Model LLM Machine learning Natural Language Processing https://urss.warwick.ac.uk/files/original/68cb57120e5c1569b58b91abaaba5609.pdf 570660eefb9326c9bb8d6d40ad573470 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Science, Engineering and Medicine Academic Poster An academic poster Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Assessing the Accuracy of Machine-Learning Models for Chemical Reaction Prediction Subject The topic of the resource Chemistry Creator An entity primarily responsible for making the resource Rastislav Turányi Date A point or period of time associated with an event in the lifecycle of the resource 2023 Chemistry Computational Chemistry DFT Machine learning NEB Nudged Elastic Band Rastislav Turányi Reaction Discovery