While Multimodal Large Language Models (MLLMs) have advanced GUI navigation agents, current approaches face limitations in cross-domain generalization and effective history utilization. We present a reasoning-enhanced framework that systematically integrates structured reasoning, action prediction, and history summarization. The structured reasoning component generates coherent Chain-of-Thought analyses combining progress estimation and decision reasoning, which inform both immediate action predictions and compact history summaries for future steps. Based on this framework, we train a GUI agent, \textbf{GUI-Rise}, through supervised fine-tuning on pseudo-labeled trajectories and reinforcement learning with Group Relative Policy Optimization (GRPO). This framework employs specialized rewards, including a history-aware objective, directly linking summary quality to subsequent action performance. Comprehensive evaluations on standard benchmarks demonstrate state-of-the-art results under identical training data conditions, with particularly strong performance in out-of-domain scenarios. These findings validate our framework's ability to maintain robust reasoning and generalization across diverse GUI navigation tasks. Code is available at https://leon022.github.io/GUI-Rise.
Guy Tennenholtz, Jihwan Jeong, Chih-Wei Hsu
et al.
Effective decision making in partially observable environments requires compressing long interaction histories into informative representations. We introduce Descriptive History Representations (DHRs): sufficient statistics characterized by their capacity to answer relevant questions about past interactions and potential future outcomes. DHRs focus on capturing the information necessary to address task-relevant queries, providing a structured way to summarize a history for optimal control. We propose a multi-agent learning framework, involving representation, decision, and question-asking components, optimized using a joint objective that balances reward maximization with the representation's ability to answer informative questions. This yields representations that capture the salient historical details and predictive structures needed for effective decision making. We validate our approach on user modeling tasks with public movie and shopping datasets, generating interpretable textual user profiles which serve as sufficient statistics for predicting preference-driven behavior of users.
Modern out-of-order CPUs heavily rely on speculative execution for performance optimization, with branch prediction serving as a cornerstone to minimize stalls and maximize efficiency. Whenever shared branch prediction resources lack proper isolation and sanitization methods, they may originate security vulnerabilities that expose sensitive data across different software contexts. This paper examines the fundamental components of modern Branch Prediction Units (BPUs) and investigates how resource sharing and contention affect two widely implemented but underdocumented features: Bias-Free Branch Prediction and Branch History Speculation. Our analysis demonstrates that these BPU features, while designed to enhance speculative execution efficiency through more accurate branch histories, can also introduce significant security risks. We show that these features can inadvertently modify the Branch History Buffer (BHB) update behavior and create new primitives that trigger malicious mis-speculations. This discovery exposes previously unknown cross-privilege attack surfaces for Branch History Injection (BHI). Based on these findings, we present three novel attack primitives: two Spectre attacks, namely Spectre-BSE and Spectre-BHS, and a cross-privilege control flow side-channel attack called BiasScope. Our research identifies corresponding patterns of vulnerable control flows and demonstrates exploitation on multiple processors. Finally, Chimera is presented: an attack demonstrator based on eBPF for a variant of Spectre-BHS that is capable of leaking kernel memory contents at 24,628 bit/s.
The dream sequence found in the Exagoge of Ezekiel (68-89) has long captured the attention of scholars who have seen it as either typical of Greek tragedy, representative of an early merkavah tradition, engaging in haggadic midrash, an investiture story, or a polemic against Enochic traditions. Classicists also have pointed to numerous parallels from Greek and Jewish literary traditions that might have informed the play. However, what has hitherto gone unnoticed is that Raguel’s interpretation of Moses’ dream conforms to a number of conventions for reporting enigmatic dreams in ancient Near Eastern literature. Of specific interest is Ezekiel’s use of polysemy and paronomasia to tie the dream to its interpretation. In the wider Near East, this hermeneutical strategy derives from divinatory practice generally, and consequently features in dream omen manuals and literary reports of dream interpretation.
The population structure within Archon apollinus (Herbst, 1798) across its range has been investigated using mtDNA COI sequences alongside five other gene regions for select samples in order to strengthen the phylogenetic backbone. Results strongly indicated the presence of two highly divergent lineages within this species: One confined to Western Turkey, Northeast Greece as well as the East Aegean islands (A. apollinus), and another in central Anatolia and the Levant. Accordingly, we recognize the Levantine lineage as a distinct species, A. bellargus Staudinger, [1892], stat. nov. Phenotypic differences between adults from western Turkey/Greece and the Levant, partly divergent flight times and significant morphological and ethological differences in caterpillars and pupae support these findings. While our central Anatolian samples were unambiguously part of the Levantine clade (A. bellargus), the life history data placed these populations closer to those in western Turkey (A. apollinus). The cause and consequences of this discordance are discussed.
All volumes of Professor Guthrie's great history of Greek philosophy have won their due acclaim. The most striking merits of Guthrie's work are his mastery of a tremendous range of ancient literature and modern scholarship, his fairness and balance of judgement and the lucidity and precision of his English prose. He has achieved clarity and comprehensiveness.
We overview the history of primordial black hole (PBH) research from the first papers around 50 years ago to the present epoch. The history may be divided into four periods, the dividing lines being marked by three key developments: inflation on the theoretical front and the detection of microlensing events by the MACHO project and gravitational waves by the LIGO/Virgo/KAGRA project on the observation front. However, they are also characterised by somewhat different focuses of research. The period 1967-1980 covered the groundbreaking work on PBH formation and evaporation. The period 1980-1996 mainly focussed on their formation, while the period 1996-2016 consolidated the work on formation but also collated the constraints on the PBH abundance. In the period 2016-2024 there was a shift of emphasis to the search for evidence for PBHs and - while opinions about the strength of the purported evidence vary - this has motivated more careful studies of some aspects of the subject. Certainly the soaring number of papers on PBHs in this last period indicates a growing interest in the topic.
Huy Nguyen, Christoph Treude, Patanamon Thongtanunam
With the exponential growth of AI tools that generate source code, understanding software has become crucial. When developers comprehend a program, they may refer to additional contexts to look for information, e.g. program documentation or historical code versions. Therefore, we argue that encoding this additional contextual information could also benefit code representation for deep learning. Recent papers incorporate contextual data (e.g. call hierarchy) into vector representation to address program comprehension problems. This motivates further studies to explore additional contexts, such as version history, to enhance models' understanding of programs. That is, insights from version history enable recognition of patterns in code evolution over time, recurring issues, and the effectiveness of past solutions. Our paper presents preliminary evidence of the potential benefit of encoding contextual information from the version history to predict code clones and perform code classification. We experiment with two representative deep learning models, ASTNN and CodeBERT, to investigate whether combining additional contexts with different aggregations may benefit downstream activities. The experimental result affirms the positive impact of combining version history into source code representation in all scenarios; however, to ensure the technique performs consistently, we need to conduct a holistic investigation on a larger code base using different combinations of contexts, aggregation, and models. Therefore, we propose a research agenda aimed at exploring various aspects of encoding additional context to improve code representation and its optimal utilisation in specific situations.
Modeling policies for sequential clinical decision-making based on observational data is useful for describing treatment practices, standardizing frequent patterns in treatment, and evaluating alternative policies. For each task, it is essential that the policy model is interpretable. Learning accurate models requires effectively capturing the state of a patient, either through sequence representation learning or carefully crafted summaries of their medical history. While recent work has favored the former, it remains a question as to how histories should best be represented for interpretable policy modeling. Focused on model fit, we systematically compare diverse approaches to summarizing patient history for interpretable modeling of clinical policies across four sequential decision-making tasks. We illustrate differences in the policies learned using various representations by breaking down evaluations by patient subgroups, critical states, and stages of treatment, highlighting challenges specific to common use cases. We find that interpretable sequence models using learned representations perform on par with black-box models across all tasks. Interpretable models using hand-crafted representations perform substantially worse when ignoring history entirely, but are made competitive by incorporating only a few aggregated and recent elements of patient history. The added benefits of using a richer representation are pronounced for subgroups and in specific use cases. This underscores the importance of evaluating policy models in the context of their intended use.
Muhammad Shihab Rashid, Jannat Ara Meem, Vagelis Hristidis
Open Retrieval Conversational Question Answering (OrConvQA) answers a question given a conversation as context and a document collection. A typical OrConvQA pipeline consists of three modules: a Retriever to retrieve relevant documents from the collection, a Reranker to rerank them given the question and the context, and a Reader to extract an answer span. The conversational turns can provide valuable context to answer the final query. State-of-the-art OrConvQA systems use the same history modeling for all three modules of the pipeline. We hypothesize this as suboptimal. Specifically, we argue that a broader context is needed in the first modules of the pipeline to not miss relevant documents, while a narrower context is needed in the last modules to identify the exact answer span. We propose NORMY, the first unsupervised non-uniform history modeling pipeline which generates the best conversational history for each module. We further propose a novel Retriever for NORMY, which employs keyphrase extraction on the conversation history, and leverages passages retrieved in previous turns as additional context. We also created a new dataset for OrConvQA, by expanding the doc2dial dataset. We implemented various state-of-the-art history modeling techniques and comprehensively evaluated them separately for each module of the pipeline on three datasets: OR-QUAC, our doc2dial extension, and ConvMix. Our extensive experiments show that NORMY outperforms the state-of-the-art in the individual modules and in the end-to-end system.
The study of cultural artifact provenance, tracing ownership and preservation, holds significant importance in archaeology and art history. Modern technology has advanced this field, yet challenges persist, including recognizing evidence from diverse sources, integrating sociocultural context, and enhancing interactive automation for comprehensive provenance analysis. In collaboration with art historians, we examined the handscroll, a traditional Chinese painting form that provides a rich source of historical data and a unique opportunity to explore history through cultural artifacts. We present a three-tiered methodology encompassing artifact, contextual, and provenance levels, designed to create a "Biography" for handscroll. Our approach incorporates the application of image processing techniques and language models to extract, validate, and augment elements within handscroll using various cultural heritage databases. To facilitate efficient analysis of non-contiguous extracted elements, we have developed a distinctive layout. Additionally, we introduce ScrollTimes, a visual analysis system tailored to support the three-tiered analysis of handscroll, allowing art historians to interactively create biographies tailored to their interests. Validated through case studies and expert interviews, our approach offers a window into history, fostering a holistic understanding of handscroll provenance and historical significance.
In the modern perception of the Ancient World the massive slave revolts loom largely. To the modern mind, infused, through education and mass media, with notions of sanctity of personal freedom and shamefulness of servitude, there is natural and immediate connection between the institution of slavery and armed, violent resistance to it. Ancient slaves were kept in obviously shameful and degrading state of bondage, therefore they revolted – they must have. In fact, however, large scale slave revolts are actually quite rare in world history and, in the case of Ancient Greece, all examples that one could point to are late and (at least superficially) marginal. If we limit our scope to Classical Greece (5th and 4th centuries BC), the slave revolt is virtually non-existent, unless we choose to widen the definition of slaves to include the helots of Sparta and the penests of Thessaly. This paper assumes that Messenian (helot) revolts are a separate (though perhaps related) phenomenon to slave revolts, and focus only on the latter. There are only three known cases of anything resembling a slave revolt (four, if we add the problematic case of the slave uprising of Drimacus, in the 3nd century BC Chios), and they seem rather minute in their scope and achievement, especially when compared to the contemporary massive slave wars of Roman Sicily and Italy. The paper argues that this absence is not an illusion, created, as one might argue, through a lack of interest or organized silence on the part of ancient authors, but the actual reflection of historical reality. Prospects of success for such endeavor were minimal, while the dangers involved were overwhelming. Specific conditions required for large scale slave uprisings were rarely met in Ancient Greece and consequently the phenomenon itself was rare.
Stefanos Hatzilazarou, Ioannis Anestis, Elias Pipinis
et al.
In the context of conservation and sustainable exploitation of neglected and underutilized plant genetic resources (NUPs), this study focused on six Cretan local endemic plants i.e., three monocots (Allium bourgeaui subsp. creticum, Allium dilatatum, Muscari spreitzenhoferi) and three dicots (Alyssum baldaccii, Campanula saxatilis subsp. saxatilis, Silene antri-jovis). We aimed at determining the ecological conditions needed for these plants to thrive based on their natural preferences which define their germination requirements and allow the development of species-specific propagation protocols. Secondly, we overviewed the potential of the targeted species for sustainable exploitation in three economic sectors (ornamental-horticultural, medicinal-cosmetic, agro-alimentary). The ecological profiles of each species were constructed using Geographic Information System and climate data from WorldClim. Four temperatures were examined in seed germination trials (10, 15, 20, 25oC) and germination percentages (GP) were calculated. Seed germination of monocots showed preference in more cold temperatures (70.0%, 40.0% and 71.25% at 10oC for A. bourgeaui subsp. creticum, A. dilatatum and M. spreitzenhoferi, respectively) while in two of the dicots it exhibited a wider temperature range (83.75-86.25% at 10, 15, 20oC for A. baldaccii and 90-98.75% at all temperatures tested for S. antri-jovis) while in C. saxatilis subsp. saxatilis at lower temperatures (85% and 71.25% at 10 and 15oC, respectively). The assessed taxa showed interesting value mainly for the ornamental and agro-alimentary sectors, and their potential is discussed herein in detail (first-time for A. baldaccii). Exploiting all the above results, we re-evaluated the feasibility and readiness timescale for sustainable exploitation in three economic sectors for the targeted NUPs and their upgraded assessments are first-time presented herein in detail.
The history of meta-learning methods based on gradient descent is reviewed, focusing primarily on methods that adapt step-size (learning rate) meta-parameters.
Ioannis Liritzis, Niki Evelpidou, Ilias Fikos
et al.
The Kastrouli Late Bronze settlement in Phocis province, central Greece, has been proved to have been an important center in the periphery of the Mycenaean palaces. It was reused at least partially and was cultivated until the 20th century. The presence of a flat area off the Kastrouli hill and the seasonal flooding nowadays led to the present investigation, questioning the formation of an ancient lake or marsh/swamp. A methodological approach was applied combining the digital elevation model (DEM) and GIS of the wider and confined area, examining slopes between 0 and 5 degrees (0 and 8.75%), with electrical resistivity tomography (ERT) traverses of around 300 and 500 m, reaching a depth of 100 m. The ERT data were rapidly collected on profiles and provided a cross-sectional (2D) plot. It was found that, in the area, there is a basin with a length of 100 m and a depth of around 40–50 m. The sedimentation process over the millennia has filled the basin, with the upper 5–6 m surface layers of the area having a low resistivity. The presence of two natural sinkholes with apparent engineered hydraulic works is noted to conform to drainage and produce a habitable environment, protecting the cultivated land and avoiding a swamp associated with health issues.
BACKGROUND There is an increasing demand for high quality subnational estimates of under-five mortality. In low and middle income countries, where the burden of under-five mortality is concentrated, vital registration is often lacking and household surveys, which provide full birth history data, are often the most reliable source. Unfortunately, these data are spatially sparse and so data are pulled from other sources to increase the available information. Summary birth histories represent a large fraction of the available data, and provide numbers of births and deaths aggregated over time, along with the mother's age. OBJECTIVE Specialized methods are needed to leverage this information, and previously the Brass method, and variants, have been used. We wish to develop a model-based approach that can propagate errors, and make the most efficient use of the data. Further, we strive to provide a method that does not have large computational overhead. CONTRIBUTION We describe a computationally efficient model-based approach which allows summary birth history and full birth history data to be combined into analyses of under-five mortality in a natural way. The method is based on fertility and mortality models that allow direct smoothing over time and space, with the possibility for including relevant covariates that are associated with fertility and/or mortality. We first examine the behavior of the approach on simulated data, before applying the model to survey and census data from Malawi.
We establish sharp asymptotically optimal strategies for the problem of online prediction with history dependent experts. The prediction problem is played (in part) over a discrete graph called the $d$ dimensional de Bruijn graph, where $d$ is the number of days of history used by the experts. Previous work [11] established $O(\varepsilon)$ optimal strategies for $n=2$ experts and $d\leq 4$ days of history, while [10] established $O(\varepsilon^{1/3})$ optimal strategies for all $n\geq 2$ and all $d\geq 1$, where the game is played for $N$ steps and $\varepsilon=N^{-1/2}$. In this paper, we show that the optimality conditions over the de Bruijn graph correspond to a graph Poisson equation, and we establish $O(\varepsilon)$ optimal strategies for all values of $n$ and $d$.
Lucretius sometimes speaks of the mind ‘projecting’ itself, echoing the Epicurean Greek technical term epibolē. The way in which he and other first-century BCE Epicureans use this concept, however, elevates it beyond anything we can find in Epicurus, and applies it in particular to the kind of superhuman intellectual leap of which the school’s founder was himself considered the master. This has implications for Lucretius’ relation to contemporary Epicureanism.