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 9 of 9 for “"MNLI"”.

  1. Comparing Methods for Automatic Identification of Mislabeled Data

    … Multi-Genre Natural Language Inference corpus (MNLI) as the dataset that all methods will inspect for mislabeled examples (Williams et al. [2018]). We choose RoBERTa-large (Liu et al. [2019]) as the model which generates information about MNLI to be used as input for each method, and we compare …

    washington Repository record for Comparing Methods for Automatic Identification of Mislabeled Data (opens in a new tab)

  2. Automated short-answer grading and misconception detection using large language models

    … e.g. the Multi-Genre Natural Language Inference (MNLI) corpus. The model's adaptability across unseen questions and domains on the minority class, coupled with its narrowed performance gap from unseen answers, highlights the profound impact of transfer learning on grading diverse student …

    unf Repository record for Automated short-answer grading and misconception detection using large language models (opens in a new tab) Download PDF for Automated short-answer grading and misconception detection using large language models (opens in a new tab)

  3. HALLUCINATION OR FACT: PERCEIVED VERACITY OF LARGE LANGUAGE MODELS

    … A natural language inference (NLI), DeBERTa-v2 (MNLI fine-tuned), is used to assess semantic relationships across model responses. This analysis examines how frequently each model explicitly admits uncertainty when faced with ambiguous or unverifiable prompts. Findings reveal distinct behavioral …

    nps Repository record for HALLUCINATION OR FACT: PERCEIVED VERACITY OF LARGE LANGUAGE MODELS (opens in a new tab)

  4. Towards Green AI: Evaluating Energy Efficiency for Quantized LLM Inference

    … 1B, Qwen 2.5 7B, and Qwen 2.5 14B across GLUE MNLI, MMLU, HumanEval, and GSM8K datasets, evaluating 10 post-training quantization (PTQ) strategies on NVIDIA H100, A6000, A100, and L4 GPUs. We identify a novel trend showing that quantization techniques tend to exhibit peak energy efficiency …

    waterloo Repository record for Towards Green AI: Evaluating Energy Efficiency for Quantized LLM Inference (opens in a new tab)

  5. Synthesis and characterisation of novel metal pyrophosphates

    … of MnH\(_2\)P\(_2\)O\(_7\) and a composition of MnLi\(_0\)\(_.\)\(_5\)H\(_1\)\(_.\)\(_5\)P\(_2\)O\(_7\). A metastable polymorph of TiP2O7 has been successfully synthesised at low temperature and adopts a structure similar to the ideal cubic structure. Incorporation of vanadium resulted in an …

    birmingham Download PDF for Synthesis and characterisation of novel metal pyrophosphates (opens in a new tab)

  6. The development of nuclease-mediated signal amplification technology for analyte detection

    … availability of functional recognition sites for MnlI and TspDTI. The switches became cleavable when partial recognition sites were completed by Driver Fragment (DF) molecules. During development, novel properties of REases were explored such as cleavage with recognition sites containing nicks and …

    unsw Repository record for The development of nuclease-mediated signal amplification technology for analyte detection (opens in a new tab)

  7. Unstructured to Actionable: Extracting wind event impact data for enhanced infrastructure resilience

    … on the multi-nominal language inference (MNLI) dataset. The model identifies the impact sentences by evaluating entailment probabilities with user-defined impact keywords. This method addresses the absence of manually labeled data and establishes a framework applicable to various reports. …

    vt Repository record for Unstructured to Actionable: Extracting wind event impact data for enhanced infrastructure resilience (opens in a new tab)

  8. StoryNet: A 5W1H-based knowledge graph to connect stories

    … score (average of CoLA, SST-2, MRPC, QQP, STS-B, MNLI, QNLI, and RTE) along with BLEU and ROUGE scores. A few approaches are studied for training-based analysis - using BERT, Roberta, XLNet, ALBERT, ELECTRA, and AllenNLP Transformer QA with the datasets - CVE, NVD, SQuAD v1.1, and SQuAD v2.0, and …

    umkc Repository record for StoryNet: A 5W1H-based knowledge graph to connect stories (opens in a new tab)

  9. Herbicide Injury Detection in Cotton Using Unmanned Aerial System-based Multispectral Imagery

    … at 6 days after application (DAA). The GSAVI, MNLI, MSAVI2, RDVI, and SAVI were effective at distinguishing between the untreated, the five low 2,4-D rates, and the 1X rate at 15 DAA with or without the soil background at all spatial resolutions. At 21, 28, and 35 DAA, all VIs were able to …

    ttu Repository record for Herbicide Injury Detection in Cotton Using Unmanned Aerial System-based Multispectral Imagery (opens in a new tab)