Corrections: Nov 2024 (#4110)

* Paper Metadata: 2024.emnlp-tutorials.4, closes #4109.
* Paper Metadata: {2024.acl-long.843}, closes #4101.
* Paper Revision: 2024.findings-acl.670, closes #4099.
* Paper Revision: 2022.nlp4dh-1.11, closes #4098.
* Paper Revision: 2022.wat-1.3, closes #4097.
* Paper Revision: {2024.sigdial-1.59}, closes #4095.
* Paper Revision{2024.emnlp-demo.9}, closes #4088.
* Paper Revision {2024.findings-emnlp.931}, closes #4084.
* Paper Revision: 2024.sighan-1.16, closes #4083.
* Paper Revision{2024.fever-1.3}, closes #4073.
* Paper Revision: {2020.acl-main.99}, closes #4072.
* Paper Revision{2024.wmt-1.26}, closes #4066.
* Paper Revision 2024.findings-emnlp.460, closes #4064.
* Paper Revision{2024.wmt-1.76}, closes #4063.
* Paper Revision{2024.findings-emnlp.444}, closes #4061.
* Paper Revision{2024.findings-emnlp.152}, closes #4060.
* Paper Revision{2024.customnlp4u-1.17}, closes #4055.
* Paper Revision: {2024.findings-emnlp.600}, closes #4087.
* Paper Revision 2024.conll-1.6, closes #4044.
* Paper Revision{2024.emnlp-main.270}, closes #4042.
* Paper Revision{2024.acl-long.618}, closes #4040.
* Paper Revision{2024.findings-emnlp.88}, closes #4036.
* Remove software and data for paper.

Author: anthology-assist

data/xml/2020.acl.xml

@@ -1442,13 +1442,14 @@
       <author><first>Xiao-Ming</first><last>Wu</last></author>
       <pages>1050–1060</pages>
       <abstract>User intent classification plays a vital role in dialogue systems. Since user intent may frequently change over time in many realistic scenarios, unknown (new) intent detection has become an essential problem, where the study has just begun. This paper proposes a semantic-enhanced Gaussian mixture model (SEG) for unknown intent detection. In particular, we model utterance embeddings with a Gaussian mixture distribution and inject dynamic class semantic information into Gaussian means, which enables learning more class-concentrated embeddings that help to facilitate downstream outlier detection. Coupled with a density-based outlier detection algorithm, SEG achieves competitive results on three real task-oriented dialogue datasets in two languages for unknown intent detection. On top of that, we propose to integrate SEG as an unknown intent identifier into existing generalized zero-shot intent classification models to improve their performance. A case study on a state-of-the-art method, ReCapsNet, shows that SEG can push the classification performance to a significantly higher level.</abstract>
-      <url hash="62999b32">2020.acl-main.99</url>
+      <url hash="119b1748">2020.acl-main.99</url>
       <doi>10.18653/v1/2020.acl-main.99</doi>
       <video href="http://slideslive.com/38929387"/>
       <bibkey>yan-etal-2020-unknown</bibkey>
       <revision id="1" href="2020.acl-main.99v1" hash="46ae257d"/>
       <revision id="2" href="2020.acl-main.99v2" hash="62999b32" date="2024-10-06">Change of author ordering.</revision>
       <pwccode url="https://github.com/fanolabs/0shot-classification" additional="false">fanolabs/0shot-classification</pwccode>
+      <revision id="3" href="2020.acl-main.99v3" hash="119b1748" date="2024-11-29">Minor update.</revision>
     </paper>
     <paper id="100">
       <title>Expertise Style Transfer: A New Task Towards Better Communication between Experts and Laymen</title>

data/xml/2022.nlp4dh.xml

@@ -131,11 +131,12 @@
       <author><first>Yutaka</first><last>Matsuo</last></author>
       <pages>79–84</pages>
       <abstract>Amis is an endangered language indigenous to Taiwan with limited data available for computational processing. We thus present an Amis-Mandarin dataset containing a parallel corpus of 5,751 Amis and Mandarin sentences and a dictionary of 7,800 Amis words and phrases with their definitions in Mandarin. Using our dataset, we also established a baseline for machine translation between Amis and Mandarin in both directions. Our dataset can be found at <url>https://github.com/francisdzheng/amis-mandarin</url>.</abstract>
-      <url hash="dc3f6fd4">2022.nlp4dh-1.11</url>
+      <url hash="41593d58">2022.nlp4dh-1.11</url>
       <bibkey>zheng-etal-2022-parallel</bibkey>
       <revision id="1" href="2022.nlp4dh-1.11v1" hash="a92f2c0d"/>
       <revision id="2" href="2022.nlp4dh-1.11v2" hash="dc3f6fd4" date="2024-03-10">Corrected the citation for mBART in section 4.1.2.</revision>
       <doi>10.18653/v1/2022.nlp4dh-1.11</doi>
+      <revision id="3" href="2022.nlp4dh-1.11v3" hash="41593d58" date="2024-11-29">This revision adds the page numbering and footer present in the original but missing in a previous revision.</revision>
     </paper>
     <paper id="12">
       <title>Machines in the media: semantic change in the lexicon of mechanization in 19th-century <fixed-case>B</fixed-case>ritish newspapers</title>

data/xml/2022.wat.xml

@@ -54,12 +54,13 @@
       <author><first>Yutaka</first><last>Matsuo</last></author>
       <pages>44–50</pages>
       <abstract>Jejueo is a critically endangered language spoken on Jeju Island and is closely related to but mutually unintelligible with Korean. Parallel data between Jejueo and Korean is scarce, and translation between the two languages requires more attention, as current neural machine translation systems typically rely on large amounts of parallel training data. While low-resource machine translation has been shown to benefit from using additional monolingual data during the pretraining process, not as much research has been done on how to select languages other than the source and target languages for use during pretraining. We show that using large amounts of Korean and Japanese data during the pretraining process improves translation by 2.16 BLEU points for translation in the Jejueo → Korean direction and 1.34 BLEU points for translation in the Korean → Jejueo direction compared to the baseline.</abstract>
-      <url hash="622ca468">2022.wat-1.3</url>
+      <url hash="408d0d78">2022.wat-1.3</url>
       <bibkey>zheng-etal-2022-improving</bibkey>
       <revision id="1" href="2022.wat-1.3v1" hash="dbf7f901"/>
       <revision id="2" href="2022.wat-1.3v2" hash="51ef8db4" date="2022-10-26">Fixed typos.</revision>
       <revision id="3" href="2022.wat-1.3v3" hash="622ca468" date="2024-03-10">Corrected the citation used for mBART from liu-etal-2020-multilingual to liu-etal-2020-multilingual-denoising.</revision>
       <pwcdataset url="https://paperswithcode.com/dataset/jit-dataset">JIT Dataset</pwcdataset>
+      <revision id="4" href="2022.wat-1.3v4" hash="408d0d78" date="2024-11-29">This revision adds the page numbering and footer present in the original but missing in a previous revision.</revision>
     </paper>
     <paper id="4">
       <title><fixed-case>TMU</fixed-case> <fixed-case>NMT</fixed-case> System with Automatic Post-Editing by Multi-Source <fixed-case>L</fixed-case>evenshtein Transformer for the Restricted Translation Task of <fixed-case>WAT</fixed-case> 2022</title>

data/xml/2024.acl.xml

@@ -8642,9 +8642,11 @@
       <author><first>Ashutosh</first><last>Modi</last><affiliation>IIT Kanpur</affiliation></author>
       <pages>11460-11499</pages>
       <abstract>Legal systems worldwide are inundated with exponential growth in cases and documents. There is an imminent need to develop NLP and ML techniques for automatically processing and understanding legal documents to streamline the legal system. However, evaluating and comparing various NLP models designed specifically for the legal domain is challenging. This paper addresses this challenge by proposing : Benchmark for Indian Legal Text Understanding and Reasoning. contains monolingual (English, Hindi) and multi-lingual (9 Indian languages) domain-specific tasks that address different aspects of the legal system from the point of view of understanding and reasoning over Indian legal documents. We present baseline models (including LLM-based) for each task, outlining the gap between models and the ground truth. To foster further research in the legal domain, we create a leaderboard (available at: https://exploration-lab.github.io/IL-TUR/ ) where the research community can upload and compare legal text understanding systems.</abstract>
-      <url hash="7de7ce82">2024.acl-long.618</url>
+      <url hash="7fe016c0">2024.acl-long.618</url>
       <bibkey>joshi-etal-2024-il</bibkey>
       <doi>10.18653/v1/2024.acl-long.618</doi>
+      <revision id="1" href="2024.acl-long.618v1" hash="7de7ce82"/>
+      <revision id="2" href="2024.acl-long.618v2" hash="7fe016c0" date="2024-12-01">Added Acknowledgement Section (at the end of Appendix, on Page 39).</revision>
     </paper>
     <paper id="619">
       <title><fixed-case>J</fixed-case>ump<fixed-case>C</fixed-case>oder: Go Beyond Autoregressive Coder via Online Modification</title>
@@ -11829,7 +11831,7 @@
     </paper>
     <paper id="843">
       <title><fixed-case>I</fixed-case>ndic<fixed-case>LLMS</fixed-case>uite: A Blueprint for Creating Pre-training and Fine-Tuning Datasets for <fixed-case>I</fixed-case>ndian Languages</title>
-      <author><first>Mohammed</first><last>Khan</last><affiliation>Indian Institute of Technology, Madras, Dhirubhai Ambani Institute Of Information and Communication Technology</affiliation></author>
+      <author><first>Mohammed Safi Ur Rahman</first><last>Khan</last><affiliation>Indian Institute of Technology, Madras, Dhirubhai Ambani Institute Of Information and Communication Technology</affiliation></author>
       <author><first>Priyam</first><last>Mehta</last><affiliation>Gujarat Technological University Ahmedabad</affiliation></author>
       <author><first>Ananth</first><last>Sankar</last><affiliation>Annamalai University</affiliation></author>
       <author><first>Umashankar</first><last>Kumaravelan</last><affiliation>AI4Bharat</affiliation></author>
@@ -11840,7 +11842,7 @@
       <author><first>Anoop</first><last>Kunchukuttan</last><affiliation>Microsoft</affiliation></author>
       <author><first>Pratyush</first><last>Kumar</last><affiliation>Indian Institute of Technology Madras, Dhirubhai Ambani Institute Of Information and Communication Technology</affiliation></author>
       <author><first>Raj</first><last>Dabre</last><affiliation>National Institute of Information and Communications Technology (NICT), National Institute of Advanced Industrial Science and Technology</affiliation></author>
-      <author><first>Mitesh</first><last>Khapra</last><affiliation>Indian Institute of Technology, Madras, Dhirubhai Ambani Institute Of Information and Communication Technology</affiliation></author>
+      <author><first>Mitesh M.</first><last>Khapra</last><affiliation>Indian Institute of Technology, Madras, Dhirubhai Ambani Institute Of Information and Communication Technology</affiliation></author>
       <pages>15831-15879</pages>
       <abstract>Despite the considerable advancements in English LLMs, the progress in building comparable models for other languages has been hindered due to the scarcity of tailored resources. Our work aims to bridge this divide by introducing an expansive suite of resources specifically designed for the development of Indic LLMs, covering 22 languages, containing a total of 251B tokens and 74.8M instruction-response pairs. Recognizing the importance of both data quality and quantity, our approach combines highly curated manually verified data, unverified yet valuable data, and synthetic data. We build a clean, open-source pipeline for curating pre-training data from diverse sources, including websites, PDFs, and videos, incorporating best practices for crawling, cleaning, flagging, and deduplication. For instruction-fine tuning, we amalgamate existing Indic datasets, translate/transliterate English datasets into Indian languages, and utilize LLaMa2 and Mixtral models to create conversations grounded in articles from Indian Wikipedia and Wikihow. Additionally, we address toxicity alignment by generating toxic prompts for multiple scenarios and then generate non-toxic responses by feeding these toxic prompts to an aligned LLaMa2 model. We hope that the datasets, tools, and resources released as a part of this work will not only propel the research and development of Indic LLMs but also establish an open-source blueprint for extending such efforts to other languages.</abstract>
       <url hash="28f6a48c">2024.acl-long.843</url>

data/xml/2024.conll.xml

@@ -83,9 +83,11 @@
       <author><first>Senja</first><last>Pollak</last></author>
       <pages>56-68</pages>
       <abstract>To predict upcoming text, language models must in some cases retrieve in-context information verbatim. In this report, we investigated how the ability of language models to retrieve arbitrary in-context nouns developed during training (across time) and as language models trained on the same dataset increase in size (across scale). We then asked whether learning of in-context retrieval correlates with learning of more challenging zero-shot benchmarks. Furthermore, inspired by semantic effects in human short-term memory, we evaluated the retrieval with respect to a major semantic component of target nouns, namely whether they denote a concrete or abstract entity, as rated by humans. We show that verbatim in-context retrieval developed in a sudden transition early in the training process, after about 1% of the training tokens. This was observed across model sizes (from 14M and up to 12B parameters), and the transition occurred slightly later for the two smallest models. We further found that the development of verbatim in-context retrieval is positively correlated with the learning of zero-shot benchmarks. Around the transition point, all models showed the advantage of retrieving concrete nouns as opposed to abstract nouns. In all but two smallest models, the advantage dissipated away toward the end of training.</abstract>
-      <url hash="c5018a5a">2024.conll-1.6</url>
+      <url hash="0bc0a631">2024.conll-1.6</url>
       <bibkey>armeni-etal-2024-transformer</bibkey>
       <doi>10.18653/v1/2024.conll-1.6</doi>
+      <revision id="1" href="2024.conll-1.6v1" hash="c5018a5a"/>
+      <revision id="2" href="2024.conll-1.6v2" hash="0bc0a631" date="2024-12-01">This revision corrects a legend in Figure 1, E and a typo in the caption of Figure 5.</revision>
     </paper>
     <paper id="7">
       <title><fixed-case>E</fixed-case>dit<fixed-case>E</fixed-case>val: An Instruction-Based Benchmark for Text Improvements</title>

data/xml/2024.customnlp4u.xml

@@ -228,9 +228,11 @@
       <author><first>Jan</first><last>Niehues</last></author>
       <pages>220-236</pages>
       <abstract>With the number of scientific papers published every year growing and current large language models (LLMs) showing state-of-the-art performance on natural language processing (NLP) tasks, we ask the question if LLMs could be utilized to answer questions on scientific papers.We investigate how well state-of-the-art large language models (LLMs) can answer questions on scientific paper by experimenting with long-context versions of the LLaMA 2 model and evaluating and training on the Qasper dataset.We analyze how well the LLMs handle longer papers and questions that can only be answered by accessing information from far out paragraphs. During our experiments, we see that the performance of these LLMs drops with growing length and position of relevant information.We employ different measures from simple prompts to chain-of-thought prompts and zero-shot usage to fine-tuning with QLoRA.While we still observe a performance loss with increased context length, our measures reduce the effects of this flaw, and we can achieve <tex-math>F_{1}</tex-math> scores similar to bigger models like GPT-4.</abstract>
-      <url hash="90997a25">2024.customnlp4u-1.17</url>
+      <url hash="82b31f06">2024.customnlp4u-1.17</url>
       <bibkey>hilgert-etal-2024-evaluating</bibkey>
       <doi>10.18653/v1/2024.customnlp4u-1.17</doi>
+      <revision id="1" href="2024.customnlp4u-1.17v1" hash="90997a25"/>
+      <revision id="2" href="2024.customnlp4u-1.17v2" hash="82b31f06" date="2024-11-29">This revision fixes an error in Table 7 about the training of CoLT5 (was: zero-shot, should: fine-tuned). It also adds missing information about the funding of this work.</revision>
     </paper>
     <paper id="18">
       <title><fixed-case>H</fixed-case>y<fixed-case>PA</fixed-case>-<fixed-case>RAG</fixed-case>: A Hybrid Parameter Adaptive Retrieval-Augmented Generation System for <fixed-case>AI</fixed-case> Legal and Policy Applications</title>