Datasets:

lp
stringclasses
3 values
src
stringclasses
415 values
mt
stringlengths
1
388
ref
stringlengths
3
370
raw
int64
0
100
annotators
int64
1
1
domain
stringclasses
1 value
year
int64
2.02k
2.02k
task
stringclasses
2 values
sys_id
int64
0
34
en-zh
From the coreference piece, we find out that tracing multiple references to the same object in an image by using pronouns is also difficult for vision and language models.
,通过使用代词在图像中追踪同一对象的多个引用对于视觉和语言模型来说也很困难.
从核心推理这一块,我们发现通过使用代词来追踪对图像中同一对象的多个引用,对于视觉和语言模型来说也是困难的。
52
1
ACL
2,023
multilingual
2
en-zh
In addition, we consider the caption to be the premise, and the foil is its hypothesis.
此外,我们认为标题是前提,箔纸是假设。
此外,我们认为标题是前提,而干扰词是其假设。
55
1
ACL
2,023
multilingual
21
en-zh
For the second problem of scarce data resources, we built our RNSum dataset consisting of about eighty two thousand pieces of data by collecting data from public GitHub repositories using the GitHub API.
第二个问题是我们建立的SCA国家资源和一些沙漠,由大约八万两千个数据组成,通过使用Gitap来处理公众数据。
对于第二个数据资源稀缺的问题,我们通过使用GitHub API从公共GitHub存储库收集数据,建立了由大约八万两千条数据组成的RNSum数据集。
56
1
ACL
2,023
multilingual
22
en-zh
CAS-Multi tended to yield higher ROUGE-L than CAS-Single.
C-E-A-S匹配往往会产生比C-E-A-S单个更高的红色错误,
CAS -Multi倾向于比CAS -Single产生更高的ROUGE -L。
83
1
ACL
2,023
multilingual
2
en-zh
Where we perform operator from q_i to q_j, and such expression is actually directed.
我们执行从 kiwi 到 Q, J 的运算符, 这样的表达式实际上是定向的。因此,
我们执行从q_i到q_j的运算符,这样的表达式实际上是定向的。
74
1
ACL
2,023
multilingual
21
en-zh
The classwise extractive then abstractive summarization model consists of two neural modules.
提取阻塞中交叉一些修复模式由两个较新的模型组成,
这个先按类别抽取然后抽象总结的模型是由两个神经模块组成的。
63
1
ACL
2,023
multilingual
22
en-zh
The remaining references were matched and converted to the corresponding article ids from our corpus.
剩余的参考被匹配并转换为相应的文章想法从一个团体
其余的参考文献被匹配,并转换为我们语料库中相应的条款ID。
41
1
ACL
2,023
multilingual
22
en-zh
So, we think this sentence might be misleading the model to an incorrect predictions.
我们认为这个句子可能会误导模型以错误的预测。
我们认为这个句子可能会误导模型做出错误的预测。
83
1
ACL
2,023
multilingual
22
en-zh
Each feature represents the likelihood for each class.
每个特征代表每个类的概率。
每个特征表示每个类的可能性。
95
1
ACL
2,023
multilingual
22
en-zh
All work aims to bridge the gap between people and the law by developing an effective retrieval system for statutory articles.
该工作旨在通过开发有效的法定条款检索系统来弥合人与人之间的差距,
所有工作旨在通过开发针对法定条款的高效检索系统,来弥合公民与法律之间的鸿沟。
78
1
ACL
2,023
multilingual
22
en-zh
So there are many tasks.
很多任务.
存在很多的任务。
88
1
ACL
2,023
multilingual
2
en-zh
As knowledge base, we use Wikipedia.
我们使用维基百科即时链接
我们使用维基百科作为知识库。
80
1
ACL
2,023
multilingual
22
en-zh
But we have some extra information like seventeen fewer pictures, and Steven has eight pictures, which is totally irrelevant.
但是我们有一些额外的信息,比如少了17个桃子,Stephen有8个桃子,这完全无关紧要.
但我们有一些额外的信息,例如照片少了17张,Seventeen有8张照片,但这些信息完全无关紧要。
58
1
ACL
2,023
multilingual
2
en-zh
Data scientists analyze data and mainly focus on the manipulating the data's existing features.
科学家分析数据,主要专注于操纵数据的现有特征.
数据科学家对数据进行分析,主要侧重于对数据的现有特征进行操作。
86
1
ACL
2,023
multilingual
2
en-zh
And then ah FeSTE executes entity linking phase.
然后快速执行到链接阶段。
然后,FeSTE执行实体链接阶段。
53
1
ACL
2,023
multilingual
21
en-zh
So, once we get the quantity representations, we can start to do inference.
一旦我们得到了数量表示, 我们就可以开始做推断.在
一旦我们得到数量陈述,我们就可以开始做推理。
97
1
ACL
2,023
multilingual
2
en-zh
The output texts can be divided into classwise segments based on special class-specific endpoint symbols.
出文本可以根据特殊的类特定的终点符号划分为全类的部分
输出的文本可以根据特殊的类特定端点符号分为分类段落。
71
1
ACL
2,023
multilingual
2
en-zh
Each of them has been labeled by experienced jurists with references to relevant articles from a corpus of more than twenty-two thousand six hundred legal articles from Belgian codes of law.
每一条都由经验丰富的法学家标记,并参考了比利时法典226,000多项法律条文中的相关条文。
每个问题都由经验丰富的法学家标记,并参考了比利时法典中超过22,600篇法律条款的语料库中的相关条款。
92
1
ACL
2,023
multilingual
21
en-zh
Our experiments show that vision and language models identify named objects and their presence in images well, as shown by the existence piece, but struggle to ground their interdependence and relationships in visual scenes when forced to respect linguistic indicators.
我们的实验表明,视觉和语言模型可以识别被命名的对象在图像中的存在,正如存在作品所表明的那样。但他们努力在视觉场景中建立相互依赖和关系,他们强迫尊重神秘指标。
我们的实验表明,视觉和语言模型能很好地识别命名对象及其在图片中的存在(如“存在性”部分所示),但在被迫尊重语言指标时,却很难在视觉场景中建立它们的相互依存性和关系。
75
1
ACL
2,023
multilingual
22
en-zh
Furthermore, complicated operators can be actually decomposed into these basic operators.
此外,复杂的运算符实际上可以分解成这些基本的运算符.
此外,复杂的运算符实际上可以分解成这些基本运算符。
90
1
ACL
2,023
multilingual
2
en-zh
As a sanity check, and because it's an interesting experiment, we also benchmark two text only models, GPT one and GPT two, to assess whether VALSE is solvable by these unimodal models by computing the perplexity of the correct and the foiled caption, no image here, and predicting the entry with the lowest perplexity.
一个理性检查,因为这是一个有趣的实验,我们也比较了两个文本模型。一个和两个来评估阀门是否可以通过计算正确和错误的标签没有图像的混乱,并预测输入与最低混乱
作为理智的检查,同时也因为这是一个有趣的实验,我们还对两个纯文本模型(GPT one和GPT two)进行了基准测试,以评估VALSE是否可由这些单模态模型解决,方法是抛开图像不管,计算正确标题和受干扰标题的困惑度,并预测具有最低困惑度的条目。
68
1
ACL
2,023
multilingual
22
en-zh
So in our datasets we are also given the mathematical expression which leads to the ah to this particular answer as well.
在我们的数据集中, 我们还得到了一个数学表达式, 它也导致了这个特定的答案。因此,
在我们的数据集中,还向我们提供了数学表达式,该表达式也导向这个特定的答案。
75
1
ACL
2,023
multilingual
21
en-zh
So, we think this sentence might be misleading the model to an incorrect predictions.
我们认为这个句子可能具有误导性, 是一个错误预测的模型。
我们认为这个句子可能会误导模型做出错误的预测。
72
1
ACL
2,023
multilingual
21
en-zh
We experiment with the standard TF-IDF and BM25 ranking functions.
我们实验了标准的 T、F、IDF 和 BM 2.5 排名函数。
我们使用标准的 TF-IDF和BM25排序函数进行实验。
70
1
ACL
2,023
multilingual
21
en-zh
As knowledge base, we use Wikipedia.
我们使用维基百科。
我们使用维基百科作为知识库。
80
1
ACL
2,023
multilingual
21
en-zh
Each one of these models pushes state-of-the-art on vision and language tasks such as visual question answering, visual common sense reasoning, image retrieval, phrase grounding.
这些模型中的每一个都推动了十八个视觉和语言任务的状态,如视觉问题回答视觉常识推理图像检索短语定位
这些模型中的每一个都在视觉和语言任务上推动了最先进的技术,如视觉问题回答、视觉常识推理、图像检索,以及短语领域。
50
1
ACL
2,023
multilingual
22
en-zh
So, we can actually correlate this expression with the sentence here. Alright.
我们实际上可以将这个表达式与这句话相关联.
我们实际上可以将这个表达式与这里的句子相关联。好。
80
1
ACL
2,023
multilingual
2
en-zh
To test this automatically, we apply natural language inference with the following rationale.
为了自动验证这一点,我们应用自然语言的推断,用下面的理性
为了自动检验这一点,我们应用了自然语言推理,其基本原理如下。
73
1
ACL
2,023
multilingual
22
en-zh
So, the expression is represented by e i j o p.
表达式被 E, I, J, O, P 表示, 在这里
表达式由e i j o p表示。
74
1
ACL
2,023
multilingual
21
en-zh
These were set based on previous research and other factors.
这些标准是基于先前的研究和其他实践而制定的.
这些都是基于之前的研究和其他因素设定的。
100
1
ACL
2,023
multilingual
2
en-zh
Instead, it's a structured collection of legal provisions that have a whole meaning only when considered in the overall context, that is, together with the supplementary information from the neighboring articles, the fields and subfields they belong to, and their place in the structure of the law.
相反,它是一个法律条款的结构集合,只有在考虑它们的整体上下文时才具有完整的意义,即与它们的邻近条款的补充信息,它们所属的领域和子领域以及它们在法律结构中的位置一起.
相反,它是法律条款的结构化集合,只有在整体上下文中考虑时才具有完整的意义,也就是说,连同相邻条款的补充信息,它们所属的字段和子字段,以及它们在法律结构中的位置。
90
1
ACL
2,023
multilingual
2
en-zh
But how do we test whether the vision and language models have captured this phenomena?
但是我们如何测试视觉和语言模型是否捕捉到这些现象?
但是,我们如何测试视觉和语言模型是否捕获了这种现象?
92
1
ACL
2,023
multilingual
2
en-zh
So, previous work in math word problem solving ah actually can ah be categorized into sequence to sequence and sequence to tree model.
之前在元问题解决方面的工作实际上可以分为序列对序列和序列对树模型.
之前解决数学文字问题的工作实际上可以分为序列到序列和序列到树模型。
50
1
ACL
2,023
multilingual
2
en-zh
This system has a simple learning based text classification model and outputs one of five labels such as features or bug fixes for each input commit message.
该系统有一个简单的基于学习的文本分类模型,并为每个输入提交消息输出五个标签之一,例如功能或错误修复.
这个系统有一个简单的基于学习的文本分类模型,并为每个输入的提交信息输出五个标签之一,如特征或错误修复。
85
1
ACL
2,023
multilingual
2
en-zh
For example, um, in these questions we are asking how many apples does Jake have?
例如,在这些问题中,我们问, Jake有几个苹果?
例如,在这些问题中,我们问Jake有多少个苹果?
100
1
ACL
2,023
multilingual
21
en-zh
For example, if this expression is not allowed, we can simply remove this expression in our search space.
例如如果这个表达式不允许,我们可以简单地在我们的搜索空间中删除这个表达式。
例如,如果此表达式不被允许,那么我们可以简单地在我们的搜索空间中删除此表达式。
72
1
ACL
2,023
multilingual
22
en-zh
So here if we look at this expression, eight times three plus three is actually generated twice, but in fact we should reuse the results.
如果我们看看这个表达式八次三次 proscre实际上产生了两次,但事实上我们应该减少结果
如果我们看一下这个表达式,八乘三加三实际上生成两次,但实际上我们应该重复使用结果。
62
1
ACL
2,023
multilingual
22
en-zh
So we try to revise the sentence to be something like the number of pear trees are thirty five fewer than the apple trees.
我们试图修改这句话的意思, 比如说梨树的数量比果树少了35棵.所以
我们尝试将句子修改为类似梨树的数量比苹果树少35棵。
85
1
ACL
2,023
multilingual
2
en-zh
We experiment with mean, max and CLS pooling as well as product and cosine for computing similarities.
我们尝试了平均max和cls池化,以及点积和co符号来计算相似性。以下是使用上
我们试验了均值、最大值和CLS集合,以及计算相似性的乘积和余弦。
72
1
ACL
2,023
multilingual
21
en-zh
We have also formulated a task of entering commit messages and summarizing them so that it is applicable to all projects written in English.
我们还制定了输入委员会信息和总结信息的任务,以便适用于所有项目。用英语写的
我们还制定了一项输入提交信息并对其进行总结的任务,以便适用于所有用英语写的项目。
50
1
ACL
2,023
multilingual
21
en-zh
Note that for training, we experiment with the two flavors of the bi-encoder architecture.
为了训练,我们试验了银幕架构的两种味道
请注意,针对训练,我们尝试使用了双编码器架构的两种风格。
65
1
ACL
2,023
multilingual
22
en-zh
So here we show an examples where multi-step reasoning is helpful.
在这里,我们展示了 多步推理有帮助的例子。
在这里,我们展示了一个多步骤推理有帮助的例子。
59
1
ACL
2,023
multilingual
21
en-zh
Although BM25 underperformed the trained bi-encoder significantly, its performance indicated that it's still a strong baseline for domain specific retrieval.
尽管BM25的性能明显低于受过训练的b编码器, 但它的性能表明它仍然是域特定检索的强大基准.
虽然BM25的表现明显低于训练好的双编码器,但它的表现表明,它仍然是特定领域检索的一个强大基线。
58
1
ACL
2,023
multilingual
2
en-zh
And we can easily incorporate some prior knowledge as constraint which can help improve the performance.
我们可以很容易地将一些先前的知识作为约束, 这可以帮助提高性能.
我们可以简单地将一些先前的知识作为约束,这样就可以帮助提高表现。
92
1
ACL
2,023
multilingual
2
en-zh
But the labels are not always consistent with each repository.
但幼虫并不总是与每个幼虫种子一致的。
但是,标签并不总是与每个存储库一致。
30
1
ACL
2,023
multilingual
21
en-zh
The BLEU is penalized when the system outputs a short sentence.
表是传输和系统是短暂的句子这 panay lit
当系统输出一个短句时,BLEU会受到惩罚。
30
1
ACL
2,023
multilingual
22
en-zh
Dataset fed into FeSTE.
ne tuning 方
数据集送入FeSTE。
4
1
ACL
2,023
multilingual
22
en-zh
Overall, the median number of citations for these cited articles is two, and less than twenty-five percent of them are cited more than five times.
总的来说,这些被引用的文章的鲸类中位数是两个, 只有不到25%被引用超过五次。
总体而言,这些被引用条款的引用次数中位数为2次,其中引用次数超过5次的不到25%。
50
1
ACL
2,023
multilingual
21
en-zh
Here are the results for our experiments.
这是我们实验的结果。
以下是我们实验的结果。
98
1
ACL
2,023
multilingual
21
en-zh
To do so, we partner with the Belgian law firm that receives each year around four thousand emails from Belgian citizens who ask for advice on a personal legal issue.
所以我们与比利时律师事务所合作,每年接收来自比利时公民的约4000名男性,他们在个人法律问题上寻求咨询
为此,我们与比利时律师事务所合作,该事务所每年收到约4000封来自比利时公民的电子邮件,他们就个人法律问题征求意见。
61
1
ACL
2,023
multilingual
22
en-zh
The second method, method, which we call CAS-Multi, consists of four different seq2seq networks, each of which correspond to one of the fixed release note classes.
细分基于特殊的卡斯特和点我们呼吁的第二种方法是由四个不同的网络组成,每个相应的车辆不是车库。
第二种方法,方法,我们称之为CAS-Multi,由四个不同的seq2seq网络组成,每个网络都对应于一个固定的发行说明类别。
10
1
ACL
2,023
multilingual
22
en-zh
Do vision and language models focus on the right thing?
视觉和语言模型是否专注于正确的事情,或
视觉和语言模型关注的是正确的事情吗?
86
1
ACL
2,023
multilingual
22
en-zh
But if we give some more reasoning description, the model is able to predict the reasoning description and also make a correct prediction here.
但是如果我们给出更多的推理描述, 模型就可以预测推理描述, 并做出正确的预测。
但是,如果我们给出更多的推理描述,那么这里的模型将能够预测推理描述,同样也会做出正确的预测。
89
1
ACL
2,023
multilingual
21
en-zh
First, the corpus of article is limited to those collected from the thirty two considered Belgian codes, which does not cover the entire Belgian law as articles from decrees, directives and ordinances are missing.
首先,条文库仅限于从32个被认为是比利时法典中收集的条文,而这并不涵盖整个比利时法律,因为法令、指令和条例中没有条文。
首先,条款的内容仅限于从比利时32部法典中收集的条款,这并不包括整个比利时的法律,因为法令、指令和条例中的条款都没有。
73
1
ACL
2,023
multilingual
21
en-zh
This is quite similar to relation extraction.
这与关系提取非常相似.
这与关系提取 非常相似。
80
1
ACL
2,023
multilingual
2
en-zh
And it also allows us to impose certain constraints from prior from prior knowledge.
它还允许我们从事事先知识的某些限制.因此
它还允许我们根据先前的知识施加某些约束。
33
1
ACL
2,023
multilingual
2
en-zh
And this is the ah algorithm for our fine, reformulated finetuning approach.
这是我们重新设计的微调方法的算法.
而这就是我们精细的、重新制定的微调方法的算法。
88
1
ACL
2,023
multilingual
2
en-zh
But the drawbacks are the performance is actually generally not better than the structured model and its lack of interpretability for prediction.
但缺点在于性能通常不比结构模型好, 并且缺乏可解释性.
但缺点是,其表现实际上并不比结构化模型好,并且缺乏用于预测的可解释性。
85
1
ACL
2,023
multilingual
2
en-zh
This information thus implies that the answer contained in the remaining relevant articles might be incomplete, although it's still completely appropriate.
这些信息并不意味着其余相关文章中的答案可能不完整,尽管它仍然是完全合适的。
因此,这一信息意味着其余相关条款中包含的答案可能是不完整的,尽管它仍然完全合适。
52
1
ACL
2,023
multilingual
22
en-zh
Besides, statutory law is not a stack of independent articles that can be treated as a complete source of information on their own, unlike news or recipes, for example.
此外,法定法律并不是一堆独立的文章,它们本身可以被视为完整的信息来源。例如,不像新闻或食谱,
此外,成文法不是一堆可以自身作为完整信息来源的独立条款,举例来说,不像新闻或食谱那样。
90
1
ACL
2,023
multilingual
21
en-zh
Here we show an example of q_1 to obtain the representation for q_2 divided by q_2 and then times q_3.
这里我们展示了一个Q1的示例,以获得Q1的表示,除以Q2,然后乘以Q3.接下来
这里我们展示了一个q_1的例子,以获得针对q_2除以q_2再乘以q_3的表达。
75
1
ACL
2,023
multilingual
2
en-zh
So this figure is taken from the PaLM paper where they perform prompting to solve the network problem in the few shot learning scenario.
这个数字是从皮尔论文中提取的,他们在几次简短的学习中提出解决金属问题的问题。
这个数字取自PaLM的论文,他们在这个论文中进行了提示,以解决少样本学习情况下的网络问题。
55
1
ACL
2,023
multilingual
22
en-zh
To shed more light on this aspect, we propose a more task agnostic direction and introduce VALSE that tests the sensitivity of vision and language models to specific linguistic phenomena that affect both the linguistic and the visual modalities.
为了更多地阐明这一方面,我们提出了一个更具任务的无神论的方向。并引入测试视觉和语言模型的敏感性对影响语言和视觉模式的特定语言现象的阀门
为了进一步阐明这方面的问题,我们提出了一个与任务更加无关的方向,并引入VALSE,对于影响语言和视觉形态的特定语言现象,其测试视觉和语言模型的敏感性。
50
1
ACL
2,023
multilingual
22
en-zh
Foiling basically means that we take the caption of an image and produce a foil by altering the caption such that it does not describe the image anymore.
通过修改标题,使其不再描述图像,我们通过专注于6个特定
干扰的意思是,我们获取一个图像的标题,通过改变标题,使其不再描述图像中的物体,从而造成干扰。
79
1
ACL
2,023
multilingual
2
en-zh
Counting and coreference also are pieces that have more than one instrument.
计数和共参考也是具有多个乐器的乐器。
计数和共指也是具有多种工具的方面。
91
1
ACL
2,023
multilingual
21
en-zh
We have built a new dataset for automatic release note generation.
我们已经建立了一个新的日集 自动生成的原因。
我们已经建立了一个新的数据集,用于自动生成发行说明。
75
1
ACL
2,023
multilingual
21
en-zh
Hence, some question are better suited than others to the statutory article retrieval task, and the domain of the less suitable ones remains to be determined.
因此,一些问题比其他人更适合法定文章检索任务,而不太合适的领域仍需确定。
因此,有些问题比其他问题更适合于法定条款检索任务,而不太适合的问题所在的领域还有待确定。
57
1
ACL
2,023
multilingual
22
en-zh
So on the left hand side, we can see if we give some examples with just question and answers, we might not be able to obtain the correct answers.
在左边,我们可以看到如果我们只是用问题和答案给出一些例子, 我们可能无法得到正确的答案.
在左侧我们可以看到,如果我们给出一些只有问题和答案的例子,我们可能无法获得正确的答案。
73
1
ACL
2,023
multilingual
2
en-zh
And here we briefly show the results compared with the previous best approaches.
在这里我们简要地展示了与以前最好的方法相比的结果.所以
在这里,我们简要地展示了与之前最佳方法相比的结果。
94
1
ACL
2,023
multilingual
2
en-zh
Dataset fed into FeSTE.
数据集被输入快速,
数据集送入FeSTE。
50
1
ACL
2,023
multilingual
2
en-zh
These results indicate that CEAS and CAS are significantly affected.
这些结果表明这一点,并且非常有效。
这些结果表明CEAS和CAS受到严重影响。
76
1
ACL
2,023
multilingual
22
en-zh
But it adds reformulation phase.
但它增加了一个“再”表示阶段,
但它增加了重新制定阶段。
95
1
ACL
2,023
multilingual
21
en-zh
Instead, the landlord should probably rely more on case law and find precedents similar to their current situation.
相反,房东可能应该更多地依赖法律,并找到类似于他们目前的情况的先例,
相反,房东可能应该更多地依靠判例法,找到与他们目前情况相似的先例。
74
1
ACL
2,023
multilingual
22
en-zh
This penalty results in a lower BLEU value in the experiment results described next.
这种忏悔导致了一个较低的自治市镇。在实验结果中, 下面描述。
在接下来描述的实验结果中,这种惩罚导致了较低的BLEU值。
6
1
ACL
2,023
multilingual
21
en-zh
In this example to finetune the language model, to classify ah to classify text into classes, abstract into classes, low or high.
节为目标数据集,以找到对语言模型进行分类以将文本分类为抽象类到低ohio类。
在这个例子中,要对语言模型进行微调,将文本分类,抽象成类,低或高。
61
1
ACL
2,023
multilingual
22
en-zh
Given a query and an article, a lexical model assigns a score to the query article pair by computing the sum over the query terms of the weights of each of these terms in that article.
给定一个文章中的查询,词汇模型通过计算查询术语的总和,在该文章中的每个术语的权重,为查询-文章对分配分数.
给定一个查询和一个条款,词汇模型通过计算该条款中每个术语的权重总和来为查询条款对分配一个分数。
77
1
ACL
2,023
multilingual
2
en-zh
It's notable how instruments centered on the individual objects like existence and noun phrases are almost solved by ViLBERT twelve in one, highlighting that models are capable of identifying named objects and their presence in images.
值得注意的是,以单个物体为中心的仪器, 比如存在和名词短语, 几乎被威尔伯特的 12 in One 解决了, 凸显出模型能够识别 有名字的物体 和它们在图像中的存在。
值得注意的是,以个别对象(如存在性和名词短语)为中心的工具几乎被ViLBERT十二分之一解决了,这突出表明模型能够识别已命名的对象和它们在图片中的存在。
85
1
ACL
2,023
multilingual
21
en-zh
In particular, on the clean test set, the score gap between the proposed method and the baselines jumped to more than twenty points.
特别是在清理的数据集上,拟议方法与基线之间的分数差距跃升至20多个点.
特别是在干净的测试集上,建议的方法和基线之间的分数差距跃升到20分以上。
73
1
ACL
2,023
multilingual
2
en-zh
Lastly, we passed the legal references and filtered out the questions whose references were not articles in one of the codes of law we considered.
最后,我们分析了法律参考文献,并过濾了那些参考文献不是我们所考虑的法律法典中的条款的问题.
最后,我们浏览了法律参考文献,并过滤掉参考文献不包含在我们所考虑的其中任何一部法律条款的问题。
70
1
ACL
2,023
multilingual
2
en-zh
Using all datasets, we benchmarked several retrieval approaches, including lexical and dense architecture.
使用数据集,我们基准了几个检索开关,包括语法和密度架构一
使用所有数据集,我们对几种检索方法进行了基准测试,包括词汇和密集架构。
45
1
ACL
2,023
multilingual
22
en-zh
We also, we also split each dataset into four folds and apply four folds cross validation.
我们还将每个数据集分成四个组, 并应用四个组的交叉验证.
我们还将每个数据集分成四折,并应用四折交叉验证。
72
1
ACL
2,023
multilingual
2
en-zh
And we only consider basic operators such as addition, subtractions, multiplication, division, and exponential.
我们只考虑加法、减法、乘法、除法和指数等基本运算符。
并且我们只考虑基本运算符,如加法、减法、乘法、除法和指数。
100
1
ACL
2,023
multilingual
21
en-zh
But in fact, in practice, in the inference stage, we might ah be able to get the incorrect expression as well.
在实践中,在推断阶段,我们也可能能够得到错误的错误的表达。在
但实际上,在实践中,在推理阶段,我们也可能会得到错误的表达式。
75
1
ACL
2,023
multilingual
22
en-zh
Okay, let me explain the experiments.
,让我解释一下实验.
好,让我解释一下实验。
75
1
ACL
2,023
multilingual
2
en-zh
And we can also see the number of all the possible ah expression is different from the previous step.
我们也可以看到所有可能的表达式的数量不同于前一步.所以
我们还可以看到,所有可能的表达式的数字与之前的步骤不同。
70
1
ACL
2,023
multilingual
2
en-zh
We assume the number of unused quantities can be regarded as irrelevant information here.
我们假定未使用的数量可被视为相关信息。
我们假设这里的未使用数量的数量可以被视为不相关的信息。
73
1
ACL
2,023
multilingual
22
en-zh
We use in our experiments base BERT base architecture.
在我们的实验中使用基于构建的基础架构
在我们的实验中,我们使用基础BERT基础架构。
65
1
ACL
2,023
multilingual
22
en-zh
So, we can actually limit this search space by removing those results that are negatives so that we can make the ah make the answer correct.
我们实际上可以通过删除那些负结果来限制这个搜索空间,这样我们就可以使答案正确.因此
我们实际上可以通过删除负数的结果来限制这个搜索空间,这样我们就可以得出正确的答案。
59
1
ACL
2,023
multilingual
2
en-zh
For example, in the case of the actions piece, we have two instruments, one in which the action verb is changed with a different action, and one in which actants are swapped.
例如,在动作部分中, 我们有两个乐器, 一个是用不同的动作 来改变动作动词, 另一个是用重音来交换。
例如,在动作方面,我们有两种工具,在一种当中是用不同的动作改变动作动词,在另一种当中是动作被交换。
50
1
ACL
2,023
multilingual
21
en-zh
Applied the language model to the new task and the output likelihood for each class.
将语言模型应用于每个类的新任务和喜欢的输出请注意
将语言模型应用于新的任务,并对每一类的输出可能性进行分析。
75
1
ACL
2,023
multilingual
22
en-zh
We hope that our work sparks interest in developing practical and reliable statutory article retrieval models.
我们希望所有的工作都激发了对开发实用和可靠的法定文章检索模型的兴趣,
我们希望我们的工作能够激发在开发实用可靠的法定条款检索模型方面的兴趣。
95
1
ACL
2,023
multilingual
22
en-zh
In this example, the original dataset has two classes.
这个例子原始数据集有两个类,
在这个例子中,原始数据集有两个类别。
94
1
ACL
2,023
multilingual
22
en-zh
Finally, we also calculate the specificity because ROUGE and BLEU cannot be calculated if the release notes are empty.
最后,我们还计算了特异性,因为如果释放说明是空的,就无法计算红色和蓝色.
最后,我们还计算了特异性,因为如果发行说明是空的,就无法计算ROUG和BLEU。
50
1
ACL
2,023
multilingual
2
en-zh
But do we know what the models have actually learned?
但我们知道模型实际学到了什么吗?
但我们是否知道模型实际上学到了什么?
85
1
ACL
2,023
multilingual
21
en-zh
We considered thirty two publicly available Belgian codes and extracted all the articles as well as the corresponding section headings.
我们考虑了32个公开可用的比利时代码,并提取了所有文章以及相应的部分标题
我们考虑了32个公开可用的比利时法典,并提取了所有条款以及相应的章节标题。
92
1
ACL
2,023
multilingual
22
en-zh
And here we briefly show the results compared with the previous best approaches.
在这里,我们简要地显示了与最好的方法相比的结果,所以
在这里,我们简要地展示了与之前最佳方法相比的结果。
63
1
ACL
2,023
multilingual
22
en-zh
But this procedure is not perfect, it is just an indicator for valid foils.
但这个程序并不完美,它只是有效的叶子的指标,
但这个过程并不完美,它只是一个有效干扰词的指标。
82
1
ACL
2,023
multilingual
22
en-zh
To overcome this limitation, we experiment with a neural based architecture that can capture semantic relationships between queries and article.
为了克服这一限制,我们实验了一种基于神经的架构,可以捕捉查询和文章之间的语义关系。
为了克服这个限制,我们尝试了一种基于神经的架构,它可以捕获查询和条款之间的语义关系。
70
1
ACL
2,023
multilingual
21
en-zh
Note that VALSE does not deliver any training data but only test data.
是,VALS不提供任何训练数据,而仅提供测试数据,
请注意,VALSE不提供任何训练数据,而仅提供测试数据。
96
1
ACL
2,023
multilingual
2
en-zh
Before diving into the main contribution of this work, let's first describe the problem of statutory article retrieval.
在沉浸在本作品的主要贡献之前,首先可以描述法定文章检索的问题。
在深入探讨这项工作的主要贡献之前,让我们首先描述法定条款检索的问题。
78
1
ACL
2,023
multilingual
22