Filters








47,054 Hits in 1.6 sec

Neural Prompt Search [article]

Yuanhan Zhang, Kaiyang Zhou, Ziwei Liu
2022 arXiv   pre-print
of prompt modules through a neural architecture search algorithm, specifically for each downstream dataset.  ...  In this paper, we view the existing parameter-efficient tuning methods as "prompt modules" and propose Neural prOmpt seArcH (NOAH), a novel approach that learns, for large vision models, the optimal design  ...  Neural Architecture Search Neural architecture search (NAS) consists of two crucial components: search space and search algorithm.  ... 
arXiv:2206.04673v2 fatcat:uneuot27d5bhvkeeu7kakopm3m

Cognitive memory

Bernard Widrow, Juan Carlos Aragon
2013 Neural Networks  
This same signal vector serves as a prompt signal for searching memory.  ...  A prompt vector from the prompt line can initiate a search in this memory segment containing folders with stored data.  ... 
doi:10.1016/j.neunet.2013.01.016 pmid:23453302 fatcat:2tvwnxinhbes3om5h5xr6wg644

Training and Inference Methods for High-Coverage Neural Machine Translation

Michael Yang, Yixin Liu, Rahul Mayuranath
2020 Proceedings of the Fourth Workshop on Neural Generation and Translation  
For inference, encouraging a small amount of diversity with Diverse Beam Search to improve translation coverage yielded marginal improvement over regular Beam Search.  ...  In this paper, we introduce a system built for the Duolingo Simultaneous Translation And Paraphrase for Language Education (STA-PLE) shared task at the 4th Workshop on Neural Generation and Translation  ...  ., 2020) at the 4th Workshop on Neural Generation and Translation (WNGT 2020). The shared task datasets consist of English prompts and a weighted set of target language translations for each prompt.  ... 
doi:10.18653/v1/2020.ngt-1.13 dblp:conf/aclnmt/YangLM20 fatcat:lio3cjlv7vhw5dvqvorbwedzue

Study of neutron tagging for Hyper-Kamiokande

M. Harada
2019 Zenodo  
Event candidate search Analysis & Evaluation Prompt signal information is used for the candidate search. →Reconstruction of n-Capture position is needed.  ...  Result of reconstruction Mis-identified Neural network Analysis result Detection efficiency in Neural network is estimated. Removed event is more than expected in (Table. 3).  ... 
doi:10.5281/zenodo.3333449 fatcat:wt2zh7cgxzdw7brorwhk5hpyqy

Improving Coherence and Consistency in Neural Sequence Models with Dual-System, Neuro-Symbolic Reasoning [article]

Maxwell Nye, Michael Henry Tessler, Joshua B. Tenenbaum, Brenden M. Lake
2021 arXiv   pre-print
Our approach uses neural inference to mediate between the neural System 1 and the logical System 2.  ...  Results in robust story generation and grounded instruction-following show that this approach can increase the coherence and accuracy of neurally-based generations.  ...  Test-time search. At test time, the neural only single-system baseline from Heinze-Deml & Bouchacourt (2020) performs greedy decoding.  ... 
arXiv:2107.02794v2 fatcat:o3gmhgj6yvcgfhh2c6l6kgnljy

Mimic and Rephrase: Reflective Listening in Open-Ended Dialogue

Justin Dieter, Tian Wang, Arun Tejasvi Chaganty, Gabor Angeli, Angel X. Chang
2019 Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL)  
We then evaluate three models for generating responses: a syntax-aware rulebased system, a seq2seq LSTM neural models with attention (S2SA), and the same neural model augmented with a copy mechanism (S2SA  ...  log p c (y i = copy(x j )|y <i , x), Modified Beam Search We use a modified version of beam search (Huang et al., 2017) when generating output tokens to favor longer responses.  ...  Neural Models To address these issues, we develop a series of neural models for the task.  ... 
doi:10.18653/v1/k19-1037 dblp:conf/conll/DieterWCAC19 fatcat:nsbvwldtavcjrkhssm3utd5z7y

Prompt-tuned Code Language Model as a Neural Knowledge Base for Type Inference in Statically-Typed Partial Code [article]

Qing Huang, Zhiqiang Yuan, Zhenchang Xing, Xiwei Xu, Liming Zhu, Qinghua Lu
2022 arXiv   pre-print
Unlike existing symbolic name and context matching for type inference, our prompt-tuned code MLM packs FQN syntax and usage in its parameters and supports fuzzy neural type inference.  ...  Built on source code naturalness, our approach fine-tunes a code masked language model (MLM) as a neural knowledge base of code elements with a novel "pre-train, prompt and predict" paradigm from raw source  ...  Subsequent work proposes to automate the search of prompts, including both discrete prompts [84, 85] and continuous prompts [86] [87] [88] .  ... 
arXiv:2208.05361v1 fatcat:um3wjyerkbg45km3jevt7ichtq

A Study of Hadronic Backgrounds to Isolated Hard Photon Production with L3 [article]

David Kirkby
1995 arXiv   pre-print
By extrapolating results obtained with L3, I estimate that the rate of prompt-photon + jet background to a H -> gamma gamma search at the LHC will be larger than Monte Carlo predictions by a factor of  ...  I describe two methods for studying hadronic backgrounds to prompt photon production with L3, and compare the observed background rates with Monte Carlo predictions.  ...  By combining these results, I estimate that the prompt photon + jets background to an LHC H! search will be larger than Monte Carlo predictions by a factor of 1.5{2.5.  ... 
arXiv:hep-ex/9505012v1 fatcat:q2zjn2vaaresbfftbdpxhi2iim

P^3 Ranker: Mitigating the Gaps between Pre-training and Ranking Fine-tuning with Prompt-based Learning and Pre-finetuning [article]

Xiaomeng Hu, Shi Yu, Chenyan Xiong, Zhenghao Liu, Zhiyuan Liu, Ge Yu
2022 arXiv   pre-print
To mitigate these gaps, we propose Pre-trained, Prompt-learned and Pre-finetuned Neural Ranker (P^3 Ranker).  ...  Compared to other language tasks, applying pre-trained language models (PLMs) for search ranking often requires more nuances and training signals.  ...  This work mitigates the gaps with a Pre-trained, Prompt-learned and Pre-finetuned Neural Ranker (P 3 Ranker).  ... 
arXiv:2205.01886v1 fatcat:esqrvmki7nec3jrwjcupptt26y

Deep Active Learning for Dialogue Generation [article]

Nabiha Asghar, Pascal Poupart, Xin Jiang, Hang Li
2017 arXiv   pre-print
We propose an online, end-to-end, neural generative conversational model for open-domain dialogue.  ...  While most existing research proposes offline supervision or hand-crafted reward functions for online reinforcement, we devise a novel interactive learning mechanism based on hamming-diverse beam search  ...  Li et al. (2016d) propose DRL-based diversity-promoting Beam Search (Koehn et al., 2003) for response generation.  ... 
arXiv:1612.03929v5 fatcat:b26h4odcvnb6bffoozvxgm2evq

Simultaneous paraphrasing and translation by fine-tuning Transformer models

Rakesh Chada
2020 Proceedings of the Fourth Workshop on Neural Generation and Translation  
This paper describes the third place submission to the shared task on simultaneous translation and paraphrasing for language education at the 4th workshop on Neural Generation and Translation (WNGT) for  ...  This has 6 encoder and 6 decoder layers and an 8-headed attention mechanism Transformer Beam Search Post-processing English Prompt N-best hypotheses Multi-output translations Figure 1 : Architecture  ...  Tan et al. (2019) train a Transformer-based Neural Machine Translation model for Hungarian-English and Portugese-English translation.  ... 
doi:10.18653/v1/2020.ngt-1.23 dblp:conf/aclnmt/Chada20 fatcat:s5v4nokv3zcgxp6yfpyujerz5e

Reconstruction of Large Radius Tracks with the Exa.TrkX pipeline [article]

Chun-Yi Wang, Xiangyang Ju, Shih-Chieh Hsu, Daniel Murnane, Paolo Calafiura, Steven Farrell, Maria Spiropulu, Jean-Roch Vlimant, Adam Aurisano, Jeremy Hewes, Giuseppe Cerati, Lindsey Gray (+13 others)
2022 arXiv   pre-print
This new capability offered by the Exa.TrkX pipeline may enable us to search for new physics in real time.  ...  Trained with all tracks in the event, the pipeline simultaneously reconstructs prompt tracks and large radius tracks with high efficiencies.  ...  Figure 2 shows the distribution of edge scores of prompt and displaced tracks obtained from graph neural network, respectively.  ... 
arXiv:2203.08800v1 fatcat:2vv233rz5vfgppgc7gsgroljqa

Deep Active Learning for Dialogue Generation

Nabiha Asghar, Pascal Poupart, Xin Jiang, Hang Li
2017 Proceedings of the 6th Joint Conference on Lexical and Computational Semantics (*SEM 2017)  
We propose an online, end-to-end, neural generative conversational model for opendomain dialogue.  ...  While most existing research proposes offline supervision or hand-crafted reward functions for online reinforcement, we devise a novel interactive learning mechanism based on hamming-diverse beam search  ...  Li et al. (2016d) propose DRL-based diversitypromoting Beam Search (Koehn et al., 2003) for response generation.  ... 
doi:10.18653/v1/s17-1008 dblp:conf/starsem/AsgharPJL17 fatcat:pfimq5yzf5bjnibxnxscqpkdsm

InPars: Data Augmentation for Information Retrieval using Large Language Models [article]

Luiz Bonifacio, Hugo Abonizio, Marzieh Fadaee, Rodrigo Nogueira
2022 arXiv   pre-print
Extensive research in various NLP tasks has shown that using domain-specific training data, as opposed to a general-purpose one, improves the performance of neural models.  ...  This makes In-Pars also suitable for non-neural retrieval algorithms.  ...  OpenAI Search API: We also experiment with OpenAI's Search API 3 as a reranker of 100 documents retrieved by BM25.  ... 
arXiv:2202.05144v1 fatcat:ghsvhhenljhjlf66e4krhu7dzi

The use of Convolutional Neural Networks for signal-background classification in Particle Physics experiments

Venkitesh Ayyar, Wahid Bhimji, Prabhat, Spencer Klein, Sally Robertson, Nick Choma, Tomasz Palczewski, Zahra Ronaghi, Lisa Gerhardt
2019 Zenodo  
The success of Convolutional Neural Networks (CNNs) in image classification has prompted efforts to study their use for classifying image data obtained in Particle Physics experiments.  ...  In this work, we present an extensive 3D convolutional neural architecture search, achieving high accuracy for signal/background discrimination for a HEP classification use-case based on simulated data  ...  signalbackground classification Prior Deep Neural Networks for Physics Analysis on low-level whole-detector data at the LHC, Bhimji et al, arxiv: 1711.03573 4.ATLAS Collaboration 2016 Search for massive  ... 
doi:10.5281/zenodo.3599134 fatcat:ckfn6ljhargd3mjm32vlb4crtq
« Previous Showing results 1 — 15 out of 47,054 results