Pub Date : 1900-01-01DOI: 10.18653/v1/2022.insights-1.22
M. Alexeeva, Allegra A. Beal, M. Surdeanu
In this paper, we introduce and justify a new task—causal link extraction based on beliefs—and do a qualitative analysis of the ability of a large language model—InstructGPT-3—to generate implicit consequences of beliefs. With the language model-generated consequences being promising, but not consistent, we propose directions of future work, including data collection, explicit consequence extraction using rule-based and language modeling-based approaches, and using explicitly stated consequences of beliefs to fine-tune or prompt the language model to produce outputs suitable for the task.
{"title":"Combining Extraction and Generation for Constructing Belief-Consequence Causal Links","authors":"M. Alexeeva, Allegra A. Beal, M. Surdeanu","doi":"10.18653/v1/2022.insights-1.22","DOIUrl":"https://doi.org/10.18653/v1/2022.insights-1.22","url":null,"abstract":"In this paper, we introduce and justify a new task—causal link extraction based on beliefs—and do a qualitative analysis of the ability of a large language model—InstructGPT-3—to generate implicit consequences of beliefs. With the language model-generated consequences being promising, but not consistent, we propose directions of future work, including data collection, explicit consequence extraction using rule-based and language modeling-based approaches, and using explicitly stated consequences of beliefs to fine-tune or prompt the language model to produce outputs suitable for the task.","PeriodicalId":441528,"journal":{"name":"First Workshop on Insights from Negative Results in NLP","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124916509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1900-01-01DOI: 10.18653/v1/2022.insights-1.3
Sopan Khosla, Rashmi Gangadharaiah
Open-world classification in dialog systems require models to detect open intents, while ensuring the quality of in-domain (ID) intent classification. In this work, we revisit methods that leverage distance-based statistics for unsupervised out-of-domain (OOD) detection. We show that despite their superior performance on threshold-independent metrics like AUROC on test-set, threshold values chosen based on the performance on a validation-set do not generalize well to the test-set, thus resulting in substantially lower performance on ID or OOD detection accuracy and F1-scores. Our analysis shows that this lack of generalizability can be successfully mitigated by setting aside a hold-out set from validation data for threshold selection (sometimes achieving relative gains as high as 100%). Extensive experiments on seven benchmark datasets show that this fix puts the performance of these methods at par with, or sometimes even better than, the current state-of-the-art OOD detection techniques.
{"title":"Evaluating the Practical Utility of Confidence-score based Techniques for Unsupervised Open-world Classification","authors":"Sopan Khosla, Rashmi Gangadharaiah","doi":"10.18653/v1/2022.insights-1.3","DOIUrl":"https://doi.org/10.18653/v1/2022.insights-1.3","url":null,"abstract":"Open-world classification in dialog systems require models to detect open intents, while ensuring the quality of in-domain (ID) intent classification. In this work, we revisit methods that leverage distance-based statistics for unsupervised out-of-domain (OOD) detection. We show that despite their superior performance on threshold-independent metrics like AUROC on test-set, threshold values chosen based on the performance on a validation-set do not generalize well to the test-set, thus resulting in substantially lower performance on ID or OOD detection accuracy and F1-scores. Our analysis shows that this lack of generalizability can be successfully mitigated by setting aside a hold-out set from validation data for threshold selection (sometimes achieving relative gains as high as 100%). Extensive experiments on seven benchmark datasets show that this fix puts the performance of these methods at par with, or sometimes even better than, the current state-of-the-art OOD detection techniques.","PeriodicalId":441528,"journal":{"name":"First Workshop on Insights from Negative Results in NLP","volume":"215 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121277892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1900-01-01DOI: 10.18653/v1/2022.insights-1.25
Philipp Koch, M. Aßenmacher, C. Heumann
Evaluating generated text received new attention with the introduction of model-based metrics in recent years. These new metrics have a higher correlation with human judgments and seemingly overcome many issues of previous n-gram based metrics from the symbolic age. In this work, we examine the recently introduced metrics BERTScore, BLEURT, NUBIA, MoverScore, and Mark-Evaluate (Petersen). We investigate their sensitivity to different types of semantic deterioration (part of speech drop and negation), word order perturbations, word drop, and the common problem of repetition. No metric showed appropriate behaviour for negation, and further none of them was overall sensitive to the other issues mentioned above.
{"title":"Pre-trained language models evaluating themselves - A comparative study","authors":"Philipp Koch, M. Aßenmacher, C. Heumann","doi":"10.18653/v1/2022.insights-1.25","DOIUrl":"https://doi.org/10.18653/v1/2022.insights-1.25","url":null,"abstract":"Evaluating generated text received new attention with the introduction of model-based metrics in recent years. These new metrics have a higher correlation with human judgments and seemingly overcome many issues of previous n-gram based metrics from the symbolic age. In this work, we examine the recently introduced metrics BERTScore, BLEURT, NUBIA, MoverScore, and Mark-Evaluate (Petersen). We investigate their sensitivity to different types of semantic deterioration (part of speech drop and negation), word order perturbations, word drop, and the common problem of repetition. No metric showed appropriate behaviour for negation, and further none of them was overall sensitive to the other issues mentioned above.","PeriodicalId":441528,"journal":{"name":"First Workshop on Insights from Negative Results in NLP","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129845894","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1900-01-01DOI: 10.18653/v1/2022.insights-1.18
Ionut-Teodor Sorodoc, Laura Aina, Gemma Boleda
To successfully account for language, computational models need to take into account both the linguistic context (the content of the utterances) and the extra-linguistic context (for instance, the participants in a dialogue). We focus on a referential task that asks models to link entity mentions in a TV show to the corresponding characters, and design an architecture that attempts to account for both kinds of context. In particular, our architecture combines a previously proposed specialized module (an “entity library”) for character representation with transfer learning from a pre-trained language model. We find that, although the model does improve linguistic contextualization, it fails to successfully integrate extra-linguistic information about the participants in the dialogue. Our work shows that it is very challenging to incorporate extra-linguistic information into pre-trained language models.
{"title":"Challenges in including extra-linguistic context in pre-trained language models","authors":"Ionut-Teodor Sorodoc, Laura Aina, Gemma Boleda","doi":"10.18653/v1/2022.insights-1.18","DOIUrl":"https://doi.org/10.18653/v1/2022.insights-1.18","url":null,"abstract":"To successfully account for language, computational models need to take into account both the linguistic context (the content of the utterances) and the extra-linguistic context (for instance, the participants in a dialogue). We focus on a referential task that asks models to link entity mentions in a TV show to the corresponding characters, and design an architecture that attempts to account for both kinds of context. In particular, our architecture combines a previously proposed specialized module (an “entity library”) for character representation with transfer learning from a pre-trained language model. We find that, although the model does improve linguistic contextualization, it fails to successfully integrate extra-linguistic information about the participants in the dialogue. Our work shows that it is very challenging to incorporate extra-linguistic information into pre-trained language models.","PeriodicalId":441528,"journal":{"name":"First Workshop on Insights from Negative Results in NLP","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130031489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：