Large language models (LLMs) have remarkable capabilities, but adapting them to specialized domains poses a fundamental question: how should accumulated experience be represented and leveraged?Existing approaches represent experience either as explicit textual artifacts in prompts (e.g., retrieved documents or dialogues) or implicitly within model weights via fine-tuning (e.g., LoRA adapters).However, textual methods are limited by context windows and cannot internalize knowledge, while parametric fine-tuning yields one adapter per task with minimal cross-task skill reuse.We propose ReX (Reusable eXperience), an experience-centric adaptation framework that treats latent experiences -recurring reasoning patterns and skills -as fundamental units for LLM specialization.Our method learns a shared Experience Bank of foundational skill vectors and uses a VAE-based encoder to map each input to a low-dimensional experience code.An Experience Router then dynamically composes the relevant skill vectors from this bank into a lightweight adapter for that input.By reusing skills across inputs, ReX enables implicit knowledge sharing across tasks without any explicit task identifiers.Experiments on multi-task NLP benchmarks show that this approach outperforms standard task-specific fine-tuning, yielding improved generalization through flexible skill reuse.Code is available at https://github. com/iLearn-Lab/ACL26-ReX.

In autonomous vehicles (AVs), especially in fully AVs, “drivers” perceive vehicle operation from a passenger’s perspective. This study focuses on the perspective change of drivers especially in fully AVs. To investigate the effect of change in perspective on drivers’ assessment of AVs, this study conducted a driving simulator experiment and a survey using different samples. The driving simulator experiment was a within-subjects design including 40 participants. The experimental results indicated that although AVs drive exactly the same as the drivers, 85% of these drivers had different assessments for AVs compared to driving themselves (self-AV bias). Among these biased drivers, 80% changed the direction of their assessments (e.g., satisfied with their own driving behaviors but dissatisfied with the same behaviors from the AVs). Additionally, self-AV bias increased with the level of self-evaluation of their own driving skills. The higher the level of self-evaluation, the higher the optimism bias (drivers think they drive better than AVs) of drivers. Thereafter, to show whether and how self-AV bias affects drivers’ intuitive acceptance of AVs, a survey was conducted, which included 381 valid questionnaires. The survey results showed that self-AV bias was a critical factor promoting perceived usefulness of AVs; particularly, highly self-AV-biased drivers responded more to the perceived usefulness of AVs. Moreover, drivers who could accept more risky driving behaviors of AVs compared to themselves also perceived AVs to be more useful. The findings from this study provide insights for understanding drivers’ assessment and acceptance of AVs’ driving behaviors.

Celebrities’ whereabouts are of pervasive importance. For instance, where politicians go, how often they visit, and who they meet, come with profound geopolitical and economic implications. Although news articles contain travel information of celebrities, it is not possible to perform large-scale and network-wise analysis due to the lack of automatic itinerary detection tools. To design such tools, we have to overcome difficulties from the heterogeneity among news articles: 1) One single article can be noisy, with irrelevant people and locations, especially when the articles are long. 2) Though it may be helpful if we consider multiple articles together to determine a particular trip, the key semantics are still scattered across different articles intertwined with various noises, making it hard to aggregate them effectively. 3) Over 20% of the articles refer to celebrity trips indirectly, instead of using the exact celebrity names or location names, leading to large portions of trips escaping regular detecting algorithms. We model text content across articles related to each candidate location as a graph to better associate essential information and cancel out the noises. Besides, we design a special pooling layer based on attention mechanism and node similarity, reducing irrelevant information from longer articles. To make up the missing information resulted from indirect mentions, we construct knowledge sub-graphs for named entities (person, organization, facility, etc.). Specifically, we dynamically update embeddings of event entities like the G7 summit from news descriptions since the properties (date and location) of the event change each time, which is not captured by pre-trained event representations. The proposed CeleTrip jointly trains these modules, which outperforms all baseline models and achieves 82.53% in the F1 metric. By open-sourcing the first tool and a carefully curated dataset for such a task, we hope to facilitate relevant research in celebrity itinerary mining as well as the social and political analysis built upon the extracted trips.

Achieving gender equality is an important pillar for humankind’s sustainable future. Pioneering data-driven gender bias research is based on large-scale public records such as scientific papers, patents, and company registrations, covering female researchers, inventors and entrepreneurs, and so on. Since gender information is often missing in relevant datasets, studies rely on tools to infer genders from names. However, available open-sourced Chinese gender-guessing tools are not yet suitable for scientific purposes, which may be partially responsible for female Chinese being underrepresented in mainstream gender bias research and affect their universality. Specifically, these tools focus on character-level information while overlooking the fact that the combinations of Chinese characters in multi-character names, as well as the components and pronunciations of characters, convey important messages. As a first effort, we design a Chinese Heterogeneous Graph Attention (CHGAT) model to capture the heterogeneity in component relationships and incorporate the pronunciations of characters. Our model largely surpasses current tools and also outperforms the state-of-the-art algorithm. Last but not least, the most popular Chinese name-gender dataset is single-character based with far less female coverage from an unreliable source, naturally hindering relevant studies. We open-source a more balanced multi-character dataset from an official source together with our code, hoping to help future research promoting gender equality.

Transportation demand forecasting is a critical precondition of optimal online transportation dispatch, which will greatly reduce drivers’ wasted mileage and customers’ waiting time, contributing to economic and environmental sustainability. Though various methods have been developed, the core spatio-temporal complexity remains challenging from three perspectives: (1) Compound spatial relationships. According to our empirical analysis, these relationships widely exist. Previous studies focus on capturing different spatial relationships using multi-homogeneous graphs. However, the information flow across various spatial relationships is not modeled explicitly. (2) Heterogeneity in spatial relationships. A region’s neighbors under the same spatial relationship may have different weights for this region. Meanwhile, different relationships may also weigh differently. (3) Synchronicity between compound spatial relationships and temporal relationships. Previous research considers synchronous influences from spatial and temporal relationships in a homogeneous fashion while compound spatial relationships are not captured for this synchronicity. To address the aforementioned perspectives, we propose the S patio- T emporal H eterogeneous graph A ttention N etwork (STHAN), where the key intuition is capturing the compound spatial relationships via meta-paths explicitly. We first construct a spatio-temporal heterogeneous graph including multiple spatial relationships and temporal relationships and use meta-paths to depict compound spatial relationships. To capture the heterogeneity, we use hierarchical attention, which contains node level attention and meta-path level attention. The synchronicity between temporal relationships and spatial relationships, including compound ones, is modeled in meta-path-level attention. Our framework outperforms state-of-the-art models by reducing 6.58%, 4.57%, and 4.20% of WMAPE in experiments on three real-world datasets, respectively.

Many recommender systems optimize click through rates (CTRs) as one of their core goals, and it further breaks down to predicting each item's click probability for a user (user-item click probability) and recommending the top ones to this particular user. User-item click probability is then estimated as a single term, and the basic assumption is that the user has different preferences over items. This is presumably true, but from real-world data, we observe that some people are naturally more active in clicking on items while some are not. This intrinsic tendency contributes to their user-item click probabilities. Besides this, when a user sees a particular item she likes, the click probability for this item increases due to this user-item preference.

In this paper, a second-order fast terminal sliding mode control (SFTSMC) scheme with backstepping is proposed to achieve the desired tracking performance for an uncertain missile's lateral system with external disturbance. This control strategy can be applied to generate lateral control commands on missile operating in flight regimes where the effectiveness of conventional aerodynamic surfaces is reduced (high angle of attack). The design procedure is divided into two steps. Firstly, we use backstepping technique to design a second-order fast terminal sliding surface (FTSS), and it drive tracking-error to converge to zero in finite time. Secondly, the SFTSMC scheme is designed by using Lyapunov's method, which can ensure the occurrence of the sliding motion in finite time. Furthermore, It can hold the character of fast transient response, improve the tracking accuracy and particularly effective to eliminate the chattering. Finally, the simulation results demonstrate the effectiveness of the proposed control scheme.