人大司法监督疏议--析人大及其常委会对审判和检察工作的监督

切实有效地实施并充分发挥人大司法监督的优越性对促进司法公正意义重大一但是，“个案监督”则不能实现人大司法监督的目的，还会导致对这一宪法原则的破坏。人大司法监督的渠道可考虑调整强化专门法律监督机关的职能，强化司法官员监察机制，注意对司法机关监督程序的监督和依法启动特定问题调查等措施得以实现：     

能动司法视阈下司法建议机制创新考量——以新形式建议创设为切入点

司法建议工作是当前各级法院坚持能动司法＼推进社会矛盾化解和社会管理创新工作的重要手段。诸多新形式建议不断创设过程中不具备必要形式要件的新形式建议书的出现表征出司法建议机制的创新过度与泛化现象。司法建议机制创新应当回归考量的四个限定原则,即立足审判原则、合法有效原则、及时必要原则、规范合理原则,并应从优化规范体系设置、强化沟通联动机制、完善内部程序以及建立人大备案评查机制等方面予以具体落实。     

论辩护律师与委托人关系的司法伦理

辩护律师与委托人之间的司法伦理经常面临着激烈的道德冲突,由于人们对此缺少认识,因此给刑辩律师的执业带来了许多压力。要想解决实践当中刑辩律师紧缺的问题,改善刑辩环境,有必要对刑辩律师与委托人之间的司法伦理进行梳理。     

Exploring the Role of ‘Gendered' Discourse Styles in Online Science Discussions

In this study, we examined whether gendered discourse styles were evidenced in online, synchronous, physics collaborative learning group discussions, and the extent to which such discourse patterns were related to the uptake of ideas within the group. We defined two discourse styles: the oppositional/direct style, theorized to be the socialized discourse pattern typically used by males, and the aligned/indirect style, theorized to be the socialized discourse pattern typically used by females. Our analysis indicates the presence of both styles in these chats and the styles were generally utilized along theorized, gendered lines. However, we also observed male use of the stereotypically ‘feminine' discourse style and female use of the stereotypically ‘masculine' discourse style. Moreover, we found no main effect for discourse style on the uptake of ideas. The findings indicate that, contrary to prior research in both face-to-face science classroom settings and online physics settings, ideas were taken up at relatively similar rates regardless of the gendered discourse style employed. Design implications of this study are discussed and suggestions for future research are made.     

抽象行政行为审查机制现状与司法审查制度的建立

当前我国抽象行政行为审查方式存在诸多缺陷,主要表现在审查范围有限、非司法性审查机制程序缺乏及行政相对方的参与权虚置.相关制度缺陷导致行政相对人权益难以获得有效救济,将抽象行政行为逐步纳入行政诉讼受案范围是完善我国抽象行政行为审查制度的根本途径.     

Evaluation in research funding agencies: Are structurally diverse teams biased against?

We analyze whether funding bodies are biased against diverse teams, which have often been linked to the production of transformative research. We develop a general framework that compares the drivers of success in the ex-ante grant decision process to the drivers of success in ex-post performance. We use our framework to systematically analyze the decisions of one of the major public funding organizations for scientific research worldwide, the UK's Engineering and Physical Sciences Research Council (EPSRC). We find that structurally diverse teams are not only penalized but are also biased against. Indeed, although teams that exhibit greater diversity in knowledge and skills, education, and/or scientific ability, are significantly less likely to obtain funding, they are generally more likely to be successful. Our mediating effects show that the evidence of a bias against diversity is weaker for teams led by prestigious researchers.     

A conceptual analysis of five measures of metacognitive monitoring

This paper describes five indices of metacognitive monitoring referred to as absolute accuracy, relative accuracy, bias, scatter, and discrimination. I provide definitions, formulae, and a discussion of the underlying construct that each of the five types of scores measures. I discuss the type of information provided by each measure and compare situations in which each measure is most appropriate. Recommendations are made for best measurement practice, as well as directions for future research. Recommendations focus on providing an operational definition of the construct being measured, selecting the most appropriate outcome measure, and using multiple measures whenever possible to triangulate findings.     

Towards understanding and mitigating unintended biases in language model-driven conversational recommendation

Conversational Recommendation Systems (CRSs) have recently started to leverage pretrained language models (LM) such as BERT for their ability to semantically interpret a wide range of preference statement variations. However, pretrained LMs are prone to intrinsic biases in their training data, which may be exacerbated by biases embedded in domain-specific language data (e.g., user reviews) used to fine-tune LMs for CRSs. We study a simple LM-driven recommendation backbone (termed LMRec) of a CRS to investigate how unintended bias — i.e., bias due to language variations such as name references or indirect indicators of sexual orientation or location that should not affect recommendations — manifests in substantially shifted price and category distributions of restaurant recommendations. For example, offhand mention of names associated with the black community substantially lowers the price distribution of recommended restaurants, while offhand mentions of common male-associated names lead to an increase in recommended alcohol-serving establishments. While these results raise red flags regarding a range of previously undocumented unintended biases that can occur in LM-driven CRSs, there is fortunately a silver lining: we show that train side masking and test side neutralization of non-preferential entities nullifies the observed biases without significantly impacting recommendation performance.     

Recognizing misogynous memes: Biased models and tricky archetypes

Warning: This paper contains examples of language and images which may be offensive.Misogyny is a form of hate against women and has been spreading exponentially through the Web, especially on social media platforms. Hateful content towards women can be conveyed not only by text but also using visual and/or audio sources or their combination, highlighting the necessity to address it from a multimodal perspective. One of the predominant forms of multimodal content against women is represented by memes, which are images characterized by pictorial content with an overlaying text introduced a posteriori. Its main aim is originally to be funny and/or ironic, making misogyny recognition in memes even more challenging. In this paper, we investigated 4 unimodal and 3 multimodal approaches to determine which source of information contributes more to the detection of misogynous memes. Moreover, a bias estimation technique is proposed to identify specific elements that compose a meme that could lead to unfair models, together with a bias mitigation strategy based on Bayesian Optimization. The proposed method is able to push the prediction probabilities towards the correct class for up to 61.43% of the cases. Finally, we identified the most challenging archetypes of memes that are still far to be properly recognized, highlighting the most relevant open research directions.