9607 research outputs found
Sort by
Mapping narratives of livable, inclusive futures
No full textThis study examines the potential of thick mapping in collecting narratives about co-design processes and engagement, particularly in the context of the New European Bauhaus (NEB) initiative. We draw on practical experiences from the EU-funded NEB project Desire. By incorporating diverse stakeholder perspectives and emphasising place-based narratives, we exemplify a thick mapping approach that synthesises complex layers, bridging current realities of ‘what is’ and envisioned futures. The research highlights the importance of a relational and nuanced understanding of built environment transformations, as well as an appreciation for mapping alternative narratives within it, advocating for inclusive engagement strategies that encompass social, context-sensitive, spatial, and future-oriented dimensions. Our interdisciplinary approach combines methods from urban design and futures to examine sites undergoing transformation. We offer a layered mapping approach that can be used to collect complex processes and amplify alternative stories of place, making invisible dynamics visible, thereby enriching narratives of more inclusive transformations. By recognising interconnections between people, places, and processes, our work highlights the importance of thick mapping in shaping alternative built environment futures, thereby enhancing the discourse on relational and non-linear approaches to capture co-design processes
The Password You Hope You Never Use: Use Cases for Duress Authentication
Full text linkWhile passwords are the most widely used method of authentication, concerns about duress attacks have emerged because attackers are motivated to gain access to our sensitive data. To mitigate this, researchers have proposed duress passwords, allowing users to signal to a system that they are under duress silently. We conduct a survey ( = 281) to investigate users’ perceptions of potential use cases for duress passwords. Our findings show that users perceive critical societal systems and institutions as use cases due to potentially high consequences of a successful duress attack. Further, they consider personal accounts, such as banking, to benefit from implementing duress passwords. Overall, our findings pave the way for future research aimed at developing solutions effective against duress attacks
CXL-Bench: Benchmarking Shared CXL Memory Access
No full textMemory access paths between a CPU core and memory are increasingly complex. Data can be placed on local- or remote-socket memory, and on local- and remote-die memory on modern multi-die CPUs, affecting memory access performance. Cache-coherent inter-device interconnects, such as Compute Express Link (CXL), allow a CPU core to perform load and store instructions to memory of a peripheral device. Such accesses incur higher access latency than accesses to local-socket memory and increase the access path complexity. For database system developers, it is important to understand the performance implications of these complex memory architectures. In this work, we present CXL-Bench, a benchmark framework for quantifying access performance for different memory access paths. CXL-Bench provides many configuration options, such as memory access patterns, the operating system’s memory abstraction, cache bypass options, and a distributed mode for setups with multiple servers accessing memory of the same device. We demonstrate the utility of CXL-Bench by quantifying memory access characteristics of two servers accessing a shared CXL 1.1 memory device. Our results show that memory accesses of one server to the device affect the access performance of another server accessing the same device. On the other hand, memory (de)allocations using CXL memory configured as a character device complete quickly, making frequent re-allocation of CXL memory feasible
“It's More Like the Planet's Assistant”: Re-imagining Smart Speaker Designs
No full textThe increasing pervasiveness of commercially available smart speakers calls for critical design research engagement to challenge traditional notions of utility and efficiency. In this paper, we report on the design of three provotypes (prototypes to provoke thought rather than refine the design process) – the Climate Change Smart Speakers (CCSS) – and the results from a one-week real-life probe study involving three participants. The CCSS are designed to make people reflect on smart speaker design and interactions using climate crisis as a catalyst, and each has a distinct persona: the Activist, the Robot, and the Voice of the Forest. Our preliminary findings indicate that CCSS are affectively engaging artifacts that prompt critical reflections on the sociocultural roles of smart speakers and our interactions with them. Based on this, we provide future directions for more diverse voice-user interfaces (VUI) in smart speaker design
Multiparty Asynchronous Session Types: A Mechanised Proof of Subject Reduction.
No full textSession types offer a type-based approach to describing the message exchange protocols between participants in communication-based systems. Initially, they were introduced in a binary setting, specifying communication patterns between two components. With the advent of multiparty session types (MPST), the typing discipline was extended to arbitrarily many components. In MPST, communication patterns are given in terms of global types, an Alice-Bob notation that gives a global view of how components interact. A central theorem of MPST is subject reduction: a well-typed system remains well-typed after reduction. The literature contains some formulations of MPST with proofs of subject reduction that have later been shown to be incorrect. In this paper, we show that the subject reduction proof of the original formulation of MPST by Honda et al. contains some flaws. Additionally, we provide a restriction to the theory and show that, for this fragment, subject reduction does indeed hold. Finally, we use subject reduction to show that well-typed processes never go wrong. All of our proofs are mechanised using the Coq proof assistant
The Complexity of Counting Small Sub-Hypergraphs
Full text linkSubgraph counting is a fundamental and well-studied problem whose computational complexity is well understood. Quite surprisingly, the hypergraph version of subgraph counting has been almost ignored. In this work, we address this gap by investigating the most basic sub-hypergraph counting problem: given a (small) hypergraph and a (large) hypergraph , compute the number of sub-hypergraphs of isomorphic to . Formally, for a family of hypergraphs, let #Sub() be the restriction of the problem to ; the induced variant #IndSub() is defined analogously. Our main contribution is a complete classification of the complexity of these problems. Assuming the Exponential Time Hypothesis, we prove that #Sub() is fixed-parameter tractable if and only if has bounded fractional co-independent edge-cover number, a novel graph parameter we introduce. Moreover, #IndSub() is fixed-parameter tractable if and only if has bounded fractional edge-cover number. Both results subsume pre-existing results for graphs as special cases. We also show that the fixed-parameter tractable cases of #Sub() and #IndSub() are unlikely to be in polynomial time, unless respectively #P = P and Graph Isomorphism P. This shows a separation with the special case of graphs, where the fixed-parameter tractable cases are known to actually be in polynomial time
What Monads Can and Cannot Do with a Few Extra Pages
Full text linkAbstract. The delay monad provides a way to introduce general recursion in type theory. To write programs that use a wide range of computational effects directly in type theory, we need to combine the delay monad with the monads of these effects. Here we present a first systematic study of such combinations.We study both the coinductive delay monad and its guarded recursive cousin, giving concrete examples of combining these with well-known computational effects. We also provide general theorems stating which algebraic effects distribute over the delay monad, and which do not. Lastly, we salvage some of the impossible cases by considering distributive laws up to weak bisimilarity.<br/
RomanLens: The Role Of Latent Romanization In Multilinguality In LLMs.
No full textLarge Language Models (LLMs) exhibit strong multilingual performance despite being predominantly trained on English-centric corpora. This raises a fundamental question: How do LLMs achieve such multilingual capabilities? Focusing on languages written in non-Roman scripts, we investigate the role of Romanization—the representation of non-Roman scripts using Roman characters—as a potential bridge in multilingual processing. Using mechanistic interpretability techniques, we analyze next-token generation and find that intermediate layers frequently represent target words in Romanized form before transitioning to native script, a phenomenon we term Latent Romanization. Further, through activation patching experiments, we demonstrate that LLMs encode semantic concepts similarly across native and Romanized scripts, suggesting a shared underlying representation. Additionally, for translation into non-Roman script languages, our findings reveal that when the target language is in Romanized form, its representations emerge earlier in the model’s layers compared to native script. These insights contribute to a deeper understanding of multilingual representation in LLMs and highlight the implicit role of Romanization in facilitating language transfer