Next Issue

25 July 2026, Volume 58 Issue 7 Previous Issue   
---
Academic Papers of the 28^th Annual Meeting of the China Association for Science and Technology
Reports of Empirical Studies


For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

Personalized alignment of large language models and its impact on moral judgment LI Chang-Jin, JIAO Liying, CHEN Zhen, XU Hengbin, WU Michael Shengtao, XU Yan 2026, 58 (7):  1237-1253.  doi: 10.3724/SP.J.1041.2026.1237 Abstract ( 74 )   HTML ( 4 )   PDF (1011KB) ( 20 )   With the advent of the era of human?machine symbiosis, the ethical dilemmas and algorithmic biases of large language models (LLMs) have sparked widespread societal concern. Consequently, guiding artificial intelligence technology toward beneficial development has emerged as a urgent and challenging imperative. This research explores the impact of personalized alignment based on the HEXACO personality model on the moral judgment of LLMs. Specifically, Study 1 examined and verified that LLMs can effectively manifest HEXACO personality traits by adhering to prompts. Study 2 explored the influence of personalized alignment on the utilitarian tendencies of LLMs, as well as the similarities and differences compared to human participants. The results indicate that personality prompts characterized by high Honesty?Humility, Agreeableness, and Conscientiousness significantly reduce the propensity of GPT-3.5, GPT-4, and ERNIE 3.5 to make utilitarian choices. Accordingly, we propose an LLM personalized alignment framework based on both the HEXACO personality model and the personality metatrait, highlighting the moral salience effect of personality dimensions within the Stability metatrait—namely, Honesty?Humility, Agreeableness, and Conscientiousness—in the personalized alignment of LLMs. This research provides a psychological foundation for the theoretical construction and technological pathways of LLM personalized alignment. Figures and Tables | References | Related Articles | Metrics

Academic Papers of the 28^th Annual Meeting of the China Association for Science and Technology

Select

The impact of anthropomorphism on perceived warmth-competence of AI and human-AI cooperation intention CUI Xunxuan, QIAO Ziteng, LIU Ning 2026, 58 (7):  1254-1278.  doi: 10.3724/SP.J.1041.2026.1254 Abstract ( 67 )   HTML ( 2 )   PDF (1060KB) ( 16 )   With the rapid development of artificial intelligence (AI), human-AI collaboration has garnered increasing attention. Although anthropomorphism is widely regarded as an important strategy to facilitate human-AI interaction, its effects are inconsistent and sometimes even counterproductive. Through six sequential studies, this research investigated the effect of anthropomorphism on human-AI collaboration intention, revealed the underlying psychological mechanism from a social cognitive perspective, and further examined the moderating roles of perceived threats (i.e., realistic threat and uniqueness threat). Studies 1a and 1b adopted a passive text-based manipulation and found that anthropomorphism indirectly promoted collaboration intention by enhancing perceived warmth. Study 2 used an active imagination manipulation and found that anthropomorphism simultaneously increased perceived warmth, perceived competence, and collaboration intention, with perceived warmth and competence serving as the mediating mechanism. Study 3a replicated the pathway through which anthropomorphism influences collaboration intention via perceived warmth and competence. Study 3b further found that even under high realistic threat, the mediating pathways of perceived warmth and competence remained significant, and realistic threat did not exert a moderating effect. Study 4 manipulated anthropomorphism using robotic appearance images and found that appearance anthropomorphism negatively affected human-AI collaboration intention, and this effect was moderated by perceived uniqueness threat. These findings help answer the critical questions of how and when to anthropomorphize in human-AI collaboration and provide theoretical references for optimizing the anthropomorphic design of AI and promoting human-AI collaboration. Figures and Tables | References | Related Articles | Metrics

Select

The impact of privacy risk perception on initial trust in autonomous vehicle: Differential responses of professionals and non-professionals SUN Yifei, LI Xiulan, DU Feng, QI Yue 2026, 58 (7):  1279-1296.  doi: 10.3724/SP.J.1041.2026.1279 Abstract ( 48 )   HTML ( 1 )   PDF (299KB) ( 16 )   Previous research has primarily focused on under-trust in autonomous driving systems. However, over-trust among ordinary consumers may likewise lead to system misuse and increased usage risks. Drawing on three empirical studies, the present research systematically examined the influence of professional background on initial trust, with a particular focus on comparing and calibrating the trust levels of ordinary consumers to be closer to those of industry professionals as domain experts. Study 1 found that non-professionals exhibited a tendency toward over-trust, and that perceived privacy risk interacted with professional background in predicting initial trust. Study 2 further manipulated privacy risk levels and revealed differential responses between professionals and non-professionals. Specifically, increased risk levels significantly heightened non-professionals’ perceived privacy risk and reduced their initial trust, whereas professionals’ initial trust was comparatively less affected by changes in risk levels. Study 3 further revealed an asymmetric response among non-professionals to privacy risk information. Under low-risk conditions, although perceived privacy risk increased significantly, initial trust showed no significant change. In contrast, under high-risk conditions, perceived privacy risk increased significantly while initial trust decreased significantly. Taken together, these findings reveal the interactive effects of professional background and perceived privacy risk on initial trust in autonomous vehicles, highlighting the differences between professionals and non-professionals in the formation of initial trust. The findings further suggest that designers of autonomous driving systems should adopt more targeted trust calibration strategies to address the differential responses of professionals and non-professionals. Figures and Tables | References | Related Articles | Metrics

Reports of Empirical Studies

Select

Estimation of point-light walker direction can be an efficient-coding and Bayesian decoding process SUN Mengying, RAN Ping, SUN Qi 2026, 58 (7):  1297-1311.  doi: 10.3724/SP.J.1041.2026.1297 Abstract ( 29 )   HTML ( 1 )   PDF (1425KB) ( 5 )   Bayesian observer models constrained by efficient encoding a have been adopted as a powerful framework for understanding the perception of various visual features. However, it remained unclear whether the computational process was effective in the point-light-walker (PLW) direction estimation. In the current study, participants performed PLW direction estimation tasks across different distribution conditions, which appeared to generate distinct short-term priors that showed some divergence from long-term prior learned in daily life. The presentation duration of each PLW stimulus was systematically varied (250 ms vs. 800 ms) to potentially modulate internal noise levels. Our results suggested that estimation biases might vary depending on both stimulus durations and short-term priors. To better understand these observations, we developed a series of Bayesian observer models incorporating efficient encoding schemes for PLW directions, with prior distributions modeled as short-term, long-term, or their weighted integration. Model comparisons indicated that the version incorporating efficient encoding with long-term priors might provide a relatively better account of the behavioral data. Hence, these findings could suggest that the estimation of PLW directions may be a cascade process involving efficient-encoding and Bayesian-decoding, offering an new perspective to understand the PLW direction perception. Figures and Tables | References | Related Articles | Metrics

Select

Serial dependence for same-category and cross-category stimuli in social attention WANG Da, YANG Zhihao, MEI Gaoxing 2026, 58 (7):  1312-1324.  doi: 10.3724/SP.J.1041.2026.1312 Abstract ( 36 )   HTML ( 1 )   PDF (880KB) ( 4 )   Serial dependence refers to an attractive bias of current perceptual judgments toward the recent past. This effect plays a crucial role in maintaining perceptual continuity in dynamic, noisy environments. The serial dependence effects have been widely found not only for low-level stimulus features (such as orientation) but also for high-level social cues (such as gaze direction) in social attention. Social attention is the ability to perceive others’ focus of attention and infer their intentions through social cues such as eye gaze. Since social attention can be conveyed by different social cues, it remains unclear whether the serial dependence effect could exist across different categories of social cues. Using an “inducer stimulus” paradigm, this study investigated same-category and cross-category serial dependence effects in social attention through two experiments (each including two sub-experiments). The results revealed that the serial dependence effect existed between same-category social cues (Experiment 1a), and this effect was modulated by emotional context (Experiment 1b). Importantly, the serial dependence effect was also present between cross-category social cues (Experiment 2a), but this effect was not observed between non-social and social cues (Experiment 2b). These findings reveal that the temporal integration of social attention was based on the processing of abstract social representation, thus providing a new perspective for understanding the dynamic integration mechanisms of social attention. Figures and Tables | References | Related Articles | Metrics

Select

The effect of social value orientation on source memory in reciprocal cooperation YANG Jingtong, JIANG Yingjie, LONG Yiting 2026, 58 (7):  1325-1342.  doi: 10.3724/SP.J.1041.2026.1325 Abstract ( 31 )   HTML ( 1 )   PDF (634KB) ( 8 )   Figures and Tables | References | Related Articles | Metrics

Select

fNIRS-guided TMS technique reveals a critical role of the left inferior parietal lobule in Chinese tone perception LIU Shujuan, TAN Lirou, QI Yun, HU Shuai, WANG Xiaojuan, YANG Jianfeng 2026, 58 (7):  1343-1356.  doi: 10.3724/SP.J.1041.2026.1343 Abstract ( 40 )   The left inferior parietal lobe (IPL) is a critical brain region for categorical perception of speech, and some studies have shown that it is involved in the categorical perception of Chinese tones. However, its specific role in tone perception remains unclear. The current study manipulated both tonal categories (intercategory vs. intracategory) and acoustic features (large vs. small acoustic intervals) in successively presented stimulus pairs to examine whether the IPL is sensitive to acoustic features or to the abstract representation of tonal categories during Chinese tone perception. Experiment 1 used functional near-infrared spectroscopy (fNIRS) to localize functional activation in the IPL during tone perception. Experiment 2 used transcranial magnetic stimulation (TMS) to virtually impair the IPL and establish a causal neural link between the IPL and tonal categorical perception. Experiment 1 showed that the dorsal IPL channel was sensitive to tonal categories, exhibiting more robust IPL activation induced by intercategory stimulation than intracategory stimulation. This region was also sensitive to acoustic intervals, with stronger activation induced by small acoustic intervals than by large ones. In Experiment 2, the dorsal and ventral IPL channels were virtually disrupted separately using TMS. When the dorsal IPL channel was inhibited, its sensitivity to tonal categories disappeared. In contrast, the ventral IPL channel and the left premotor cortex (PMC) channel were sensitive to acoustic intervals, with stronger activation induced by large acoustic intervals than by small ones. However, when TMS inhibited the ventral IPL channel, only functional compensation in the left inferior frontal gyrus channels showed sensitivity to tonal categories and induced greater intracategory activation than intercategory stimulation did. In summary, these findings demonstrate a critical role of the left dorsal IPL in the categorical perception of Chinese tone. In contrast, the ventral IPL plays an essential gating role in the dorsal stream of speech perception. The results enrich the neurophysiological model of speech perception from a tonal language perspective, elucidating the collaborative mechanism between dorsal and ventral speech streams in categorical perception. References | Related Articles | Metrics

Select

Gradient modulation of spatial metaphors for abstract concepts in L2 by embodied experience: ERP evidence YAO Zhao, LI Tiantian, ZHU Xiangru 2026, 58 (7):  1357-1369.  doi: 10.3724/SP.J.1041.2026.1357 Abstract ( 68 )   HTML ( 2 )   PDF (481KB) ( 14 )   Figures and Tables | References | Related Articles | Metrics

Select

The cognitive and neural mechanism by which situational pressure increases the efficiency of creative discovery behavior ZHENG Yilin, ZHANG Ling, HUANG Furong 2026, 58 (7):  1370-1386.  doi: 10.3724/SP.J.1041.2026.1370 Abstract ( 36 )   HTML ( 1 )   PDF (1602KB) ( 9 )   Monitoring and resolving cognitive conflict are key cognitive processes in creative discovery. However, due to individuals’ tendency to avoid cognitive effort, they may fail to resolve the cognitive conflict triggered by novel ideas, missing numerous opportunities to generate or identify novel products. To examine the mechanism by which situational pressure enhances cognitive control resources and improves the efficiency of novel product identification, Experiment 1 established a pressure condition, and Experiment 2 further included a high-pressure condition. The results showed that for novel but less useful products, both pressure and high-pressure conditions elicited a smaller N400 and a greater late positive component (LSP) compared to the non-pressure condition. For novel and highly useful products, only the high-pressure condition elicited a smaller N400 and a greater LSP relative to the non-pressure condition. Moreover, participants’ accuracy in identifying novel products was higher under both pressure and high-pressure conditions than under the non-pressure condition. These findings indicate that situational pressure increases cognitive control resources during creative discovery, promotes cognitive conflict resolution, and thereby enhances the efficiency of novel product identification. This study not only refines the mechanism of cognitive conflict monitoring and resolution in creative discovery but also provides a novel cognitive regulation pathway for training creative behavior. Figures and Tables | References | Related Articles | Metrics

Select

Processing mechanism of negative discounting behavior: Explanation based on query theory LIU Lei, LIU Wenxia, CHENG Yang, GE Chunlei, LIU Hongting, LI Yu 2026, 58 (7):  1387-1404.  doi: 10.3724/SP.J.1041.2026.1387 Abstract ( 37 )   HTML ( 1 )   PDF (686KB) ( 4 )   Negative discounting behavior in intertemporal decision-making poses a challenge to traditional decision models based on the assumption of positive discounting. Based on query theory, the present study investigated the processing mechanism of negative discounting behavior through four studies. Studies 1 and 2, which employed the aspect listing method, found that individuals who preferentially and more frequently retrieve “value-increasing” information about immediate losses are more inclined to choose immediate losses. Conversely, individuals who preferentially and more frequently retrieve “value-reducing” information about immediate losses are more inclined to choose delayed losses. Study 3 manipulated the query order to examine its causal impact on negative discounting behavior. The findings suggest that the process of ascertaining the reasons for an immediate loss initially leads to an enhancement in negative discounting. Study 4 further demonstrated that as the magnitude of loss increases, individuals are more likely to choose a delayed loss, and this effect arises because more value-reducing information about immediate loss is retrieved earlier. This study offers a novel theoretical perspective on understanding negative discounting behavior and provides empirical evidence for behavioral interventions. Figures and Tables | References | Related Articles | Metrics

Select

The personality structure and cultural conceptualization of humility in Chinese culture WANG Xin, WU Yanhong 2026, 58 (7):  1405-1427.  doi: 10.3724/SP.J.1041.2026.1405 Abstract ( 48 )   HTML ( 1 )   PDF (1098KB) ( 12 )   Humility is a traditional virtue esteemed in Chinese culture, exerting positive effects on individuals, organizations, and society. Combining qualitative research, corpus compilation, and a review of previous scales, this study constructed the structure of the humility trait within Chinese culture and developed the Chinese Humility Scale. The scale comprises five dimensions: Accurate Self-Perception, Other-Enhancement, Dialectical Thinking, Self-Effacement, and Self-Improvement. Based on this five-dimensional structure, Latent Profile Analysis was utilized, revealing five distinct profiles of humility manifestations in Chinese culture: “Humble,” “Moderately Humble,” “Overly Humble,” “Unhumble,” and “Strategically Humble.” This research clarifies the conceptual connotations of humility within Chinese culture and provides a novel measurement tool for future empirical studies, as well as for the identification and selection of well-rounded talents possessing both competence and humility. Furthermore, it plays a positive role in inheriting and promoting the traditional virtue of humility while facilitating individual development and social harmony. Figures and Tables | References | Related Articles | Metrics

Select

Fast = Sincere: The Effect of Interactive Partner Decision Time in Social Dilemmas LIU Yongfang, SUN Yue, LIAN Jinjing 2026, 58 (7):  1428-1443.  doi: 10.3724/SP.J.1041.2026.1428 Abstract ( 48 )   HTML ( 1 )   PDF (606KB) ( 25 )   In response to debates about how an interaction partner’s decision time affects interpersonal perception and cooperative behavior in social dilemmas, this study separated information about behavioral outcomes from decision-time information and examined the mechanism by which a pure decision-time signal operates. A pilot experiment determined appropriate time limits for fast and slow decisions. Experiment 1 used a “direct perception + indirect cue” paradigm to manipulate decision time and examined whether and how it influenced perceived sincerity. Experiment 2a used the same paradigm to examine whether decision time affected cooperation in social dilemmas through perceived sincerity and cooperation expectations. Experiments 2b and 2c used only “direct perception” or only an “indirect cue,” respectively, to examine the robustness of the results from Experiment 2a under weakened decision-time manipulations and to rule out perceived competence as an alternative explanation. Experiment 3 explored whether and how individuals’ social value orientation moderated the decision-time effect. The results showed that fast decisions led to higher perceived sincerity without affecting perceived competence, and promoted greater cooperative behavior through both an independent mediating effect of perceived sincerity and a sequential mediation pathway involving perceived sincerity and cooperation expectations. This effect applied to pro-social individuals but not to pro-self individuals. These findings not only confirm the independent social-signaling meaning of interaction partner decision time, but also reveal a “dual-pathway, two-stage” mechanism and boundary conditions through which it affects cooperation in social dilemmas. The findings provide new evidence for resolving theoretical debates about the interpersonal effects of decision time and offer practical implications for optimizing cooperation strategies in real-world social dilemmas by conveying sincerity through shorter response times. Figures and Tables | References | Related Articles | Metrics

Select

You only live once: The influence of temporal perspective on individual consumption tendency LI Yaoqi, CHEN Zhengren, JIANG Qiumin, LIN Meizhen 2026, 58 (7):  1444-1458.  doi: 10.3724/SP.J.1041.2026.1444 Abstract ( 37 )   HTML ( 1 )   PDF (205KB) ( 13 )   Figures and Tables | References | Related Articles | Metrics