Results 21 to 30 of about 28,651,668 (327)
DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts [PDF]
Annual Meeting of the Association for Computational Linguistics, 2021 Despite recent advances in natural language generation, it remains challenging to control attributes of generated text. We propose DExperts: Decoding-time Experts, a decoding-time method for controlled text generation that combines a pretrained language ...
Alisa Liu +6 more
semanticscholar +1 more source
KCTS: Knowledge-Constrained Tree Search Decoding with Token-Level Hallucination Detection [PDF]
Conference on Empirical Methods in Natural Language Processing, 2023 Large Language Models (LLMs) have demonstrated remarkable human-level natural language generation capabilities. However, their potential to generate misinformation, often called the hallucination problem, poses a significant risk to their deployment.
Sehyun Choi +3 more
semanticscholar +1 more source
Epsilon Sampling Rocks: Investigating Sampling Strategies for Minimum Bayes Risk Decoding for Machine Translation [PDF]
Conference on Empirical Methods in Natural Language Processing, 2023 Recent advances in machine translation (MT) have shown that Minimum Bayes Risk (MBR) decoding can be a powerful alternative to beam search decoding, especially when combined with neural-based utility functions.
Markus Freitag +2 more
semanticscholar +1 more source
Contrastive Decoding Improves Reasoning in Large Language Models [PDF]
arXiv.org, 2023 We demonstrate that Contrastive Decoding -- a simple, computationally light, and training-free text generation method proposed by Li et al 2022 -- achieves large out-of-the-box improvements over greedy decoding on a variety of reasoning tasks. Originally
Sean O'Brien, Mike Lewis
semanticscholar +1 more source
Decoding NGC 628 with radiative transfer methods [PDF]
Monthly Notices of the Royal Astronomical Society, 2022 ABSTRACT We present an axi-symmetric model for the ultraviolet (UV)-to-submillimetre (submm) images of the nearly face-on spiral galaxy NGC 628. It was calculated using a radiative transfer (RT) code, accounting for the absorption and re-emission of starlight by dust in the interstellar medium of this galaxy.
Rushton, Mark +4 more
openaire +2 more sources
Decoding algorithms for surface codes [PDF]
Quantum, 2023 Quantum technologies have the potential to solve certain computationally hard problems with polynomial or super-polynomial speedups when compared to classical methods.
Antonio deMarti iOlius +4 more
semanticscholar +1 more source
IEEE Journal of Radio Frequency Identification, 2020 A comparison of four decoding methods for a Chipless RFID system was executed by analyzing the read success rate of a frequency-position coding tag with 12-bit of capacity.
Raphael Tavares de Alencar +4 more
semanticscholar +1 more source
Speculative Decoding: Exploiting Speculative Execution for Accelerating Seq2seq Generation [PDF]
Conference on Empirical Methods in Natural Language Processing, 2022 We propose Speculative Decoding (SpecDec), for the first time ever, to formally study exploiting the idea of speculative execution to accelerate autoregressive (AR) decoding.
Heming Xia +5 more
semanticscholar +1 more source
End-to-End Dense Video Captioning with Parallel Decoding [PDF]
IEEE International Conference on Computer Vision, 2021 Dense video captioning aims to generate multiple associated captions with their temporal locations from the video. Previous methods follow a sophisticated "localizethen-describe" scheme, which heavily relies on numerous hand-crafted components.
Teng Wang +5 more
semanticscholar +1 more source
IEICE ESS Fundamentals Review, 2011 誤り訂正符号の代数的復号法の一つであるユークリッド復号法について解説する．ユークリッド復号法はBCH符号やRS 符号の強力な復号法として知られ，数多くの適用例がある．まず，1950 年代に始まった誤り訂正符号の構成法の研究と，1960 年代から1970 年代にかけての復号法の研究の歴史をたどり，ユークリッド復号法の位置付けを行う．次に，ユークリッド復号法の概要を紹介する．最後に，この復号法発見に至る筆者らの当時の環境や様子を，研究者の立場から経験談に私見を交えながら述べる．
Masao Kasahara, Shigeichi Hirasawa
openaire +2 more sources