Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 48 for “"GPT-2"”.

  1. Health-aware Food Planner: A Personalized Recipe Generation Approach Based on GPT-2

    … and personalized ingredients, 2) fine-tuning GPT-2 model to generate a new recipe given the ingredients, 3) finding and recommending the top similar recipes to the generated one. In contrast to other recipe generation models, we expand the final output to be the generated recipes in addition …

    wilfrid-laurier Repository record for Health-aware Food Planner: A Personalized Recipe Generation Approach Based on GPT-2 (opens in a new tab) Download PDF for Health-aware Food Planner: A Personalized Recipe Generation Approach Based on GPT-2 (opens in a new tab)

  2. Text summarization using transfer learnin: Extractive and abstractive summarization using BERT and GPT-2 on news and podcast data

    … BERT, while the latter uses the language model GPT-2. We operate on two types of datasets: CNN/DailyMail, a benchmarked news article dataset and Podcast, a dataset comprised of podcast episode transcripts. The results obtained using the GPT-2 on the CNN/DailyMail dataset are competitive to …

    chalmers Repository record for Text summarization using transfer learnin: Extractive and abstractive summarization using BERT and GPT-2 on news and podcast data (opens in a new tab)

  3. Umělá neuronová síť pro rekonstruování vymřelých druhů

    … Dále byl použit sekvenční transformátorový model GPT-2, který je naučen na trénovací množině lidského přirozeného jazyka. Tento model byl použitý pro zakódování DNA sekvencí do vektorové podoby, ve které byla zachycena sémantika a kontext mezi jednotlivými částmi DNA sekvence. Modely by bylo velmi …

    brno-tech Repository record for Umělá neuronová síť pro rekonstruování vymřelých druhů (opens in a new tab)

  4. Inferred Author Gender as a Variable Affecting LLM Behaviour

    … as a source of gender-influenced behaviour in GPT-2, using an exploratory approach based on interpretability techniques rather than the traditional “biased/unbiased” dichotomy. A simple zero-shot prompt reveals that GPT-2 has learned to estimate author gender in some contexts. Attribution …

    waterloo Repository record for Inferred Author Gender as a Variable Affecting LLM Behaviour (opens in a new tab)

  5. Constrained Text Generation

    … πως τα προεκπαιδευμένα Γλωσσικά Μοντέλα (π.χ. GPT-2, BART) από μόνα τους και χωρίς τροποποιήσεις δεν προσφέρουν αρκούντως καλά αποτελέσματα. Πολλές προσεγγίσεις έχουν προταθεί για να πετύχουν καλύτερα αποτελέσματα, χρησιμοποιώντας εξωτερικές πηγές και πόρους, σε συνδυασμό με τροποποιήσεις στα …

    athens Repository record for Constrained Text Generation (opens in a new tab)

  6. Le double numérique et l'identité : entre réflexion, amplification et diffraction

    … mes données personnelles, les ai remixé dans GPT-2 et diffusé le tout sous forme de messages publiés sur Twitter. Plutôt qu’une réflexion ou un miroir tendu, ce qui en a résulté est l’amplification de certains traits, l’amputation d’autres et au final une diffraction de plusieurs fragments de …

    uqam Download PDF for Le double numérique et l'identité : entre réflexion, amplification et diffraction (opens in a new tab)

  7. Dezinformacijos identifikavimas tekstuose /

    … this paper, the transformer architecture and the GPT-2 and BERT architectures based on it were analyzed. Two scientific papers were also analyzed which attempted to develop similar models that could classify true and false texts as accurately as possible. During the work, two Lithuanian data sets …

    vilnius Repository record for Dezinformacijos identifikavimas tekstuose / (opens in a new tab)

  8. The artefacts of intelligence: governing scientists' contribution to AI proliferation

    … case study follows: the ‘staged release’ of the GPT-2 language model, where more capable models were gradually released; the platform through which researchers and developers could access GPT-3, the successor to GPT-2; and a wave of new ethics regimes for AI conference publications. The study …

    oxford Repository record for The artefacts of intelligence: governing scientists' contribution to AI proliferation (opens in a new tab)

  9. Analyzing Factors Influencing Performance in LLM Inference Systems

    … and internal working mechanism of Llama 3 and GPT-2 to explore potential methods that could keep improving LLM inferences. The approach of this project includes 2 steps which are data collection and data analysis. Our results show that increasing the batch size significantly improves throughput …

    chalmers Repository record for Analyzing Factors Influencing Performance in LLM Inference Systems (opens in a new tab)

  10. Improve recurrent neural network language model for automatic speech recognition

    … ASR. With only 7.25 percent of the parameters of GPT-2 medium model, our proposed RNNLM achieved up to 38.59 percent WER relative reductions on the WSJ dataset, which are comparable to those of a fine-tuned GPT-2 medium model. On the AMI dataset our approach even outperformed the fine-tuned GPT-2 …

    missouri Repository record for Improve recurrent neural network language model for automatic speech recognition (opens in a new tab)

  11. Molecular characterisation of candidate wheat proteins associated with the development of type 1 diabetes.

    … were alanine aminotransferase 2 transaminase 2 (GPT-2, TC# 8197) and wheat storage globulin (WSG, Acc.No.AAA3469). GPT-2 and WSG have similar amino acid sequences as human heat shock protein 60 (HSP60), an autoantigen in TID. Moreover, WSG shares sequence homology with the insulin receptor …

    ottawa-retro Repository record for Molecular characterisation of candidate wheat proteins associated with the development of type 1 diabetes. (opens in a new tab)

  12. Deep learning-driven market volatility forecasting and model interpretation

    … memory retrieval architecture (LASER), and a GPT-2 model with an improved beam search algorithm (BEAM). In the first task, we observed a gap where most topic models have solely leveraged one level (grain) of the latent topic concept to represent and reconstruct documents. To bridge the gap, we …

    iastate Repository record for Deep learning-driven market volatility forecasting and model interpretation (opens in a new tab)

  13. Relevant Phrase Generation for Language Learners

    … prompt. We achieve this by first fine-tuning a GPT–2 model implemented by HuggingFace to a dataset of human-written sentences specifically tailored to language learners from Tatoeba.org. A decoding algorithm consisting of two steps was implemented. Initially, a shift to the probability …

    chalmers Repository record for Relevant Phrase Generation for Language Learners (opens in a new tab)

  14. AUDITING LARGE LANGUAGE MODELS: ACCOUNTABILITY, INTEGRITY, AND COPYRIGHT

    … (ZKP) protocol optimized for architectures like GPT-2. By utilizing constraint fusion and circuit squeeze, we significantly reduce computational overhead, enabling inference proofs in under 25 seconds on CPU hardware. Third, to safeguard data ownership and copyright compliance, we present …

    nus Repository record for AUDITING LARGE LANGUAGE MODELS: ACCOUNTABILITY, INTEGRITY, AND COPYRIGHT (opens in a new tab) Download file for AUDITING LARGE LANGUAGE MODELS: ACCOUNTABILITY, INTEGRITY, AND COPYRIGHT (opens in a new tab)

  15. Assessing the memorability of familiar vocabulary for system assigned passphrases

    … with the Generative Pre-trained Transformer 2 (GPT-2) model trained on the familiar vocabulary and are readable, pronounceable, sentence like passphrases resembling natural English sentences. Contrary to expectations, following a spaced repetition schedule, passphrases as natural English …

    uoit Repository record for Assessing the memorability of familiar vocabulary for system assigned passphrases (opens in a new tab)

  16. Generování stylizovaného lidského jazyka v dialogových systémech

    … nejmodernější předpřipravené modely BART a GPT-2 pro generování textu. Experimenty odhalily problém, že i současné nejmodernější modely trpí špatným odhadem kompromisu mezi stylem a kontextem. Jinými slovy, čím více se styl projeví v generované sekvenci, tím méně se vztahuje k tématu …

    brno-tech Repository record for Generování stylizovaného lidského jazyka v dialogových systémech (opens in a new tab)

  17. Estudo comparativo entre modelos transformers aplicados ao desenvolvimento de chatbots

    … DIET, LaBSE, BERTimbau, DistilBERT, RoBERTa, GPT, GPT-2 and XLNET applied to the development of a chatbot submitted to data captured in public services carried out by the Court of Auditors of the State of Rio Grande do Norte (TCE/RN). The study considered metrics such as accuracy and …

    brazil-ufrn Repository record for Estudo comparativo entre modelos transformers aplicados ao desenvolvimento de chatbots (opens in a new tab)

  18. Conditional generation of temporally-ordered event sequences

    … temporally into the input events compared to GPT-2 models.

    texas Repository record for Conditional generation of temporally-ordered event sequences (opens in a new tab)

  19. Study on Adversarial Robustness of Phishing Email Detection Models

    … phishing emails are generated using a fine-tuned GPT-2 model. The detection model has retrained with newly formed dataset and we have observed the accuracy and robustness of the model has not improved under black box attack methods. In our last experiment we proposed a defensive technique to …

    houston Repository record for Study on Adversarial Robustness of Phishing Email Detection Models (opens in a new tab)

Page 1 of 3