Search

Solidity compiler plays a key role in enabling the development of smart contract applications on Ethereum by governing the syntax of a domain-specific language called Solidity and performing compilation and optimization of Solidity code. The correctness of Solidity compiler is critical in fostering transparency, efficiency, and trust in industries reliant on smart contracts. However, like other software systems, Solidity compiler is prone to bugs, which may produce incorrect bytecodes on blockchain platforms, resulting in severe security concerns. As a domain-specific compiler for smart contracts, Solidity compiler differs from other compilers in many perspectives, posing unique challenges to detect its bugs. To understand the bugs in Solidity compiler and benefit future research, in this paper, we present the first systematic study on 533 Solidity compiler bugs. We carefully examined their characteristics (including symptoms, root causes, and distribution), and their triggering test cases. Our study leads to seven bug-revealing takeaways for Solidity compiler. Moreover, to study the limitations of Solidity compiler fuzzers and bring our findings into practical scenarios, we evaluate three Solidity compiler fuzzers on our constructed benchmark. The results show that these fuzzers are inefficient in detecting Solidity compiler bugs. The inefficiency arises from their failure to consider the interesting bug-inducing features, bug-related compilation flags, and test oracles

Background: Incidence of adverse outcome events rises as patients with advanced illness approach end-of-life. Exposures that tend to occur near end-of-life, e.g., use of wheelchair, oxygen therapy and palliative care, may therefore be found associated with the incidence of the adverse outcomes. We propose a strategy for time-to-event analysis to mitigate the time-varying confounding. Methods: We propose a concept of reverse time-to-death (rTTD) and its use for the time-scale in time-to-event analysis. We used data on community-based palliative care uptake (exposure) and emergency department visits (outcome) among patients with advanced cancer in Singapore to illustrate. We compare the results against that of the common practice of using time-on-study (TOS) as time-scale. Results: Graphical analysis demonstrated that cancer patients receiving palliative care had higher rate of emergency department visits than non-recipients mainly because they were closer to end-of-life, and that rTTD analysis made comparison between patients at the same time-to-death. Analysis of emergency department visits in relation to palliative care using TOS time-scale showed significant increase in hazard ratio estimate when observed time-varying covariates were omitted from statistical adjustment (change-in-estimate=0.38; 95% CI 0.15 to 0.60). There was no such change in otherwise the same analysis using rTTD (change-in-estimate=0.04; 95% CI -0.02 to 0.11), demonstrating the ability of rTTD time-scale to mitigate confounding that intensifies in relation to time-to-death. Conclusion: Use of rTTD as time-scale in time-to-event analysis provides a simple and robust approach to control time-varying confounding in studies of advanced illness, even if the confounders are unmeasured., Comment: 22 pages (including 2 tables and 2 figures)

Risk-driven behavior provides a feedback mechanism through which individuals both shape and are collectively affected by an epidemic. We introduce a general and flexible compartmental model to study the effect of heterogeneity in the population with regards to risk tolerance. The interplay between behavior and epidemiology leads to a rich set of possible epidemic dynamics. Depending on the behavioral composition of the population, we find that increasing heterogeneity in risk tolerance can either increase or decrease the epidemic size. We find that multiple waves of infection can arise due to the interplay between transmission and behavior, even without the replenishment of susceptibles. We find that increasing protective mechanisms such as the effectiveness of interventions, the number of risk-averse people in the population, and the duration of intervention usage reduces the epidemic overshoot. When the protection is pushed past a critical threshold, the epidemic dynamics enter an underdamped regime where the epidemic size exactly equals the herd immunity threshold. Lastly, we can find regimes where epidemic size does not monotonically decrease with a population that becomes increasingly risk-averse., Comment: 12 pages, 7 figures (main text)

Generic deep learning (DL) networks for image restoration like denoising and interpolation lack mathematical interpretability, require voluminous training data to tune a large parameter set, and are fragile in the face of covariate shift. To address these shortcomings, we build interpretable networks by unrolling variants of a graph-based optimization algorithm of different complexities. Specifically, for a general linear image formation model, we first formulate a convex quadratic programming (QP) problem with a new $\ell_2$-norm graph smoothness prior called gradient graph Laplacian regularizer (GGLR) that promotes piecewise planar (PWP) signal reconstruction. To solve the posed unconstrained QP problem, instead of computing a linear system solution straightforwardly, we introduce a variable number of auxiliary variables and correspondingly design a family of ADMM algorithms. We then unroll them into variable-complexity feed-forward networks, amenable to parameter tuning via back-propagation. More complex unrolled networks require more labeled data to train more parameters, but have better overall performance. The unrolled networks contain periodic insertions of a graph learning module, akin to a self-attention mechanism in a transformer architecture, to learn pairwise similarity structure inherent in data. Experimental results show that our unrolled networks perform competitively to generic DL networks in image restoration quality while using only a tiny fraction of parameters, and demonstrate improved robustness to covariate shift.

We report the observation of a coalescing compact binary with component masses $2.5\text{-}4.5~M_\odot$ and $1.2\text{-}2.0~M_\odot$ (all measurements quoted at the 90% credible level). The gravitational-wave signal GW230529_181500 was observed during the fourth observing run of the LIGO-Virgo-KAGRA detector network on 2023 May 29 by the LIGO Livingston Observatory. The primary component of the source has a mass less than $5~M_\odot$ at 99% credibility. We cannot definitively determine from gravitational-wave data alone whether either component of the source is a neutron star or a black hole. However, given existing estimates of the maximum neutron star mass, we find the most probable interpretation of the source to be the coalescence of a neutron star with a black hole that has a mass between the most massive neutron stars and the least massive black holes observed in the Galaxy. We provisionally estimate a merger rate density of $55^{+127}_{-47}~\text{Gpc}^{-3}\,\text{yr}^{-1}$ for compact binary coalescences with properties similar to the source of GW230529_181500; assuming that the source is a neutron star-black hole merger, GW230529_181500-like sources constitute about 60% of the total merger rate inferred for neutron star-black hole coalescences. The discovery of this system implies an increase in the expected rate of neutron star-black hole mergers with electromagnetic counterparts and provides further evidence for compact objects existing within the purported lower mass gap., Comment: 45 pages (10 pages author list, 13 pages main text, 1 page acknowledgements, 13 pages appendices, 8 pages bibliography), 17 figures, 16 tables. Update to match version published in The Astrophysical Journal Letters. Data products available from https://zenodo.org/records/10845779

Language models (LMs) exhibit impressive performance and generalization capabilities. However, LMs struggle with the persistent challenge of catastrophic forgetting, which undermines their long-term sustainability in continual learning (CL). Existing approaches usually address the issue by incorporating old task data or task-wise inductive bias into LMs. However, old data and accurate task information are often unavailable or costly to collect, hindering the availability of current CL approaches for LMs. To address this limitation, we introduce $\textbf{MIGU}$ ($\textbf{M}$agn$\textbf{I}$tude-based $\textbf{G}$radient $\textbf{U}$pdating for continual learning), a rehearsal-free and task-label-free method that only updates the model parameters with large magnitudes of output in LMs' linear layers. MIGU is based on our observation that the L1-normalized magnitude distribution of the output in LMs' linear layers is different when the LM models deal with different task data. By imposing this simple constraint on the gradient update process, we can leverage the inherent behaviors of LMs, thereby unlocking their innate CL abilities. Our experiments demonstrate that MIGU is universally applicable to all three LM architectures (T5, RoBERTa, and Llama2), delivering state-of-the-art or on-par performance across continual finetuning and continual pre-training settings on four CL benchmarks. For example, MIGU brings a 15.2% average accuracy improvement over conventional parameter-efficient finetuning baselines in a 15-task CL benchmark. MIGU can also seamlessly integrate with all three existing CL types to further enhance performance. Code is available at \href{https://github.com/wenyudu/MIGU}{this https URL}., Comment: preprint, 19 pages

This study proposes an approach for error correction in radiology reports, leveraging large language models (LLMs) and retrieval-augmented generation (RAG) techniques. The proposed framework employs a novel internal+external retrieval mechanism to extract relevant medical entities and relations from the report of interest and an external knowledge source. A three-stage inference process is introduced, decomposing the task into error detection, localization, and correction subtasks, which enhances the explainability and performance of the system. The effectiveness of the approach is evaluated using a benchmark dataset created by corrupting real-world radiology reports with realistic errors, guided by domain experts. Experimental results demonstrate the benefits of the proposed methods, with the combination of internal and external retrieval significantly improving the accuracy of error detection, localization, and correction across various state-of-the-art LLMs. The findings contribute to the development of more robust and reliable error correction systems for clinical documentation., Comment: v2

We explore a recently proposed effective field theory describing electromagnetically or gravitationally interacting massive particles in an expansion about their mass ratio, also known as the self-force (SF) expansion. By integrating out the deviation of the heavy particle about its inertial trajectory, we obtain an effective action whose only degrees of freedom are the lighter particle together with the photon or graviton, all propagating in a Coulomb or Schwarzschild background. The 0SF dynamics are described by the usual background field method, which at 1SF is supplemented by a "recoil operator" that encodes the wobble of the heavy particle, and similarly computable corrections appearing at 2SF and higher. Our formalism exploits the fact that the analytic expressions for classical backgrounds and particle trajectories encode dynamical information to all orders in the couplings, and from them we extract multiloop integrands for perturbative scattering. As a check, we study the two-loop classical scattering of scalar particles in electromagnetism and gravity, verifying known results. We then present new calculations for the two-loop classical scattering of dyons, and of particles interacting with an additional scalar or vector field coupling directly to the lighter particle but only gravitationally to the heavier particle., Comment: 77 pages, 10 figures

Programming models for distributed dataflow have long focused on analytical workloads that allow the runtime to dynamically place and schedule compute logic. Meanwhile, models that enable fine-grained control over placement, such as actors, make global optimization difficult. In this extended abstract, we present Suki, an embedded Rust DSL that lets developers implement streaming dataflow with explicit placement of computation. Key to this choreographic programming approach is our use of staged programming, which lets us expose a high-level Rust API while compiling local compute units into individual binaries with zero-overhead. We also explore how this approach, combined with Rust's trait system, enables a type-safe API for mapping dataflow programs to cloud computing resources.

Reducing the inference latency of large language models (LLMs) is crucial, and speculative decoding (SD) stands out as one of the most effective techniques. Rather than letting the LLM generate all tokens directly, speculative decoding employs effective proxies to predict potential outputs, which are then verified by the LLM without compromising the generation quality. Yet, deploying SD in real online LLM serving systems (with continuous batching) does not always yield improvement -- under higher request rates or low speculation accuracy, it paradoxically increases latency. Furthermore, there is no best speculation length work for all workloads under different system loads. Based on the observations, we develop a dynamic framework SmartSpec. SmartSpec dynamically determines the best speculation length for each request (from 0, i.e., no speculation, to many tokens) -- hence the associated speculative execution costs -- based on a new metric called goodput, which characterizes the current observed load of the entire system and the speculation accuracy. We show that SmartSpec consistently reduces average request latency by up to 3.2x compared to non-speculative decoding baselines across different sizes of target models, draft models, request rates, and datasets. Moreover, SmartSpec can be applied to different styles of speculative decoding, including traditional, model-based approaches as well as model-free methods like prompt lookup and tree-style decoding.

Long sequence modeling has gained broad interest as large language models (LLMs) continue to advance. Recent research has identified that a large portion of hidden states within the key-value caches of Transformer models can be discarded (also termed evicted) without affecting the perplexity performance in generating long sequences. However, we show that these methods, despite preserving perplexity performance, often drop information that is important for solving downstream tasks, a problem which we call information neglect. To address this issue, we introduce Chunked Instruction-aware State Eviction (CItruS), a novel modeling technique that integrates the attention preferences useful for a downstream task into the eviction process of hidden states. In addition, we design a method for chunked sequence processing to further improve efficiency. Our training-free method exhibits superior performance on long sequence comprehension and retrieval tasks over several strong baselines under the same memory budget, while preserving language modeling perplexity., Comment: Work in progress

A novel method to reconstruct the energy of hadronic showers in the CMS High Granularity Calorimeter (HGCAL) is presented. The HGCAL is a sampling calorimeter with very fine transverse and longitudinal granularity. The active media are silicon sensors and scintillator tiles readout by SiPMs and the absorbers are a combination of lead and Cu/CuW in the electromagnetic section, and steel in the hadronic section. The shower reconstruction method is based on graph neural networks and it makes use of a dynamic reduction network architecture. It is shown that the algorithm is able to capture and mitigate the main effects that normally hinder the reconstruction of hadronic showers using classical reconstruction methods, by compensating for fluctuations in the multiplicity, energy, and spatial distributions of the shower's constituents. The performance of the algorithm is evaluated using test beam data collected in 2018 prototype of the CMS HGCAL accompanied by a section of the CALICE AHCAL prototype. The capability of the method to mitigate the impact of energy leakage from the calorimeter is also demonstrated., Comment: Prepared for submission to JINST

The dramatic surge in the utilisation of generative artificial intelligence (GenAI) underscores the need for a secure and efficient mechanism to responsibly manage, use and disseminate multi-dimensional data generated by artificial intelligence (AI). In this paper, we propose a blockchain-based copyright traceability framework called ring oscillator-singular value decomposition (RO-SVD), which introduces decomposition computing to approximate low-rank matrices generated from hardware entropy sources and establishes an AI-generated content (AIGC) copyright traceability mechanism at the device level. By leveraging the parallelism and reconfigurability of field-programmable gate arrays (FPGAs), our framework can be easily constructed on existing AI-accelerated devices and provide a low-cost solution to emerging copyright issues of AIGC. We developed a hardware-software (HW/SW) co-design prototype based on comprehensive analysis and on-board experiments with multiple AI-applicable FPGAs. Using AI-generated images as a case study, our framework demonstrated effectiveness and emphasised customisation, unpredictability, efficiency, management and reconfigurability. To the best of our knowledge, this is the first practical hardware study discussing and implementing copyright traceability specifically for AI-generated content., Comment: Accepted on 20 May 2024 as a full paper at ASAP 2024

Previous foundation models for retinal images were pre-trained with limited disease categories and knowledge base. Here we introduce RetiZero, a vision-language foundation model that leverages knowledge from over 400 fundus diseases. To RetiZero's pre-training, we compiled 341,896 fundus images paired with text descriptions, sourced from public datasets, ophthalmic literature, and online resources, encompassing a diverse range of diseases across multiple ethnicities and countries. RetiZero exhibits superior performance in several downstream tasks, including zero-shot disease recognition, image-to-image retrieval, and internal- and cross-domain disease identification. In zero-shot scenarios, RetiZero achieves Top5 accuracy scores of 0.8430 for 15 fundus diseases and 0.7561 for 52 fundus diseases. For image retrieval, it achieves Top5 scores of 0.9500 and 0.8860 for the same disease sets, respectively. Clinical evaluations show that RetiZero's Top3 zero-shot performance surpasses the average of 19 ophthalmologists from Singapore, China and the United States. Furthermore, RetiZero significantly enhances clinicians' accuracy in diagnosing fundus disease. These findings underscore the value of integrating the RetiZero foundation model into clinical settings, where a variety of fundus diseases are encountered.

This paper describes our submission to the MEDIQA-CORR 2024 shared task for automatically detecting and correcting medical errors in clinical notes. We report results for three methods of few-shot In-Context Learning (ICL) augmented with Chain-of-Thought (CoT) and reason prompts using a large language model (LLM). In the first method, we manually analyse a subset of train and validation dataset to infer three CoT prompts by examining error types in the clinical notes. In the second method, we utilise the training dataset to prompt the LLM to deduce reasons about their correctness or incorrectness. The constructed CoTs and reasons are then augmented with ICL examples to solve the tasks of error detection, span identification, and error correction. Finally, we combine the two methods using a rule-based ensemble method. Across the three sub-tasks, our ensemble method achieves a ranking of 3rd for both sub-task 1 and 2, while securing 7th place in sub-task 3 among all submissions., Comment: accepted as NAACL workshop

Benchmarking is seen as critical to assessing progress in NLP. However, creating a benchmark involves many design decisions (e.g., which datasets to include, which metrics to use) that often rely on tacit, untested assumptions about what the benchmark is intended to measure or is actually measuring. There is currently no principled way of analyzing these decisions and how they impact the validity of the benchmark's measurements. To address this gap, we draw on evidence-centered design in educational assessments and propose Evidence-Centered Benchmark Design (ECBD), a framework which formalizes the benchmark design process into five modules. ECBD specifies the role each module plays in helping practitioners collect evidence about capabilities of interest. Specifically, each module requires benchmark designers to describe, justify, and support benchmark design choices -- e.g., clearly specifying the capabilities the benchmark aims to measure or how evidence about those capabilities is collected from model responses. To demonstrate the use of ECBD, we conduct case studies with three benchmarks: BoolQ, SuperGLUE, and HELM. Our analysis reveals common trends in benchmark design and documentation that could threaten the validity of benchmarks' measurements.

Testing is a major approach to ensuring the quality of deep learning (DL) libraries. Existing testing techniques commonly adopt differential testing to relieve the need for test oracle construction. However, these techniques are limited in finding implementations that offer the same functionality and generating diverse test inputs for differential testing. This paper introduces DLLens, a novel differential testing technique for DL library testing. Our insight is that APIs in different DL libraries are commonly designed to accomplish various computations for the same set of published DL algorithms. Although the mapping of these APIs is not often one-to-one, we observe that their computations can be mutually simulated after proper composition and adaptation. The use of these simulation counterparts facilitates differential testing for the detection of functional DL library bugs. Leveraging the insight, we propose DLLens as a novel mechanism that utilizes a large language model (LLM) to synthesize valid counterparts of DL library APIs. To generate diverse test inputs, DLLens incorporates a static analysis method aided by LLM to extract path constraints from all execution paths in each API and its counterpart's implementations. These path constraints are then used to guide the generation of diverse test inputs. We evaluate DLLens on two popular DL libraries, TensorFlow and PyTorch. Our evaluation shows that DLLens can synthesize counterparts for more than twice as many APIs found by state-of-the-art techniques on these libraries. Moreover, DLLens can extract 26.7% more constraints and detect 2.5 times as many bugs as state-of-the-art techniques. DLLens has successfully found 56 bugs in recent TensorFlow and PyTorch libraries. Among them, 41 are previously unknown, 39 of which have been confirmed by developers after reporting, and 19 of those confirmed bugs have been fixed by developers.

Embedders play a central role in machine learning, projecting any object into numerical representations that can, in turn, be leveraged to perform various downstream tasks. The evaluation of embedding models typically depends on domain-specific empirical approaches utilizing downstream tasks, primarily because of the lack of a standardized framework for comparison. However, acquiring adequately large and representative datasets for conducting these assessments is not always viable and can prove to be prohibitively expensive and time-consuming. In this paper, we present a unified approach to evaluate embedders. First, we establish theoretical foundations for comparing embedding models, drawing upon the concepts of sufficiency and informativeness. We then leverage these concepts to devise a tractable comparison criterion (information sufficiency), leading to a task-agnostic and self-supervised ranking procedure. We demonstrate experimentally that our approach aligns closely with the capability of embedding models to facilitate various downstream tasks in both natural language processing and molecular biology. This effectively offers practitioners a valuable tool for prioritizing model trials.

Scientific peer review is essential for the quality of academic publications. However, the increasing number of paper submissions to conferences has strained the reviewing process. This surge poses a burden on area chairs who have to carefully read an ever-growing volume of reviews and discern each reviewer's main arguments as part of their decision process. In this paper, we introduce \sys, a summarization method designed to offer a concise yet comprehensive overview of scholarly reviews. Unlike traditional consensus-based methods, \sys extracts both common and unique opinions from the reviews. We introduce novel uniqueness scores based on the Rational Speech Act framework to identify relevant sentences in the reviews. Our method aims to provide a pragmatic glimpse into all reviews, offering a balanced perspective on their opinions. Our experimental results with both automatic metrics and human evaluation show that \sys generates more discriminative summaries than baseline methods in terms of human evaluation while achieving comparable performance with these methods in terms of automatic metrics.

Code generation benchmarks such as HumanEval are widely adopted to evaluate LLMs' capabilities. However, after consolidating the latest 24 benchmarks, we noticed three significant imbalances. First, imbalanced programming language. 95.8% of benchmarks involve Python, while only 5 benchmarks involve Java. Second, imbalanced code granularity. Function-/statement-level benchmarks account for over 83.3% of benchmarks. Only a mere handful extends to class-/project-levels, and all are limited to Python. Third, lacking advanced features. Existing benchmarks primarily assess basic coding skills, while overlooking advanced Object-Oriented Programming (OOP) features (i.e., encapsulation, inheritance, and polymorphism). To fill these gaps, we propose JavaBench, a project-level Java benchmark that exercises OOP features. It comprises four Java projects with 389 methods in 106 Java classes. The test coverage is up to 92%, and JavaBench is attested by 282 undergraduate students, reaching a 90.93/100 average score (i.e., pass rate against the test suite), ensuring the quality of documentation, code skeleton, and tests. To better evaluate LLM's capability against JavaBench, we introduce a systematic evaluation design covering three context settings and five synthesis strategies at two granularities using three hierarchical metrics. Our extensive experiment yields several interesting findings. First, we noticed that regarding project-level Java programming, LLMs are far behind undergraduate students (no project can be correctly completed by any studied LLMs, and at most 41.17% Pass@5 in a more relaxed evaluation). Second, using method signature as prompt context may strike an ideal balance for project-level code generation. JavaBench is publicly available at https://github.com/java-bench/JavaBench.

Social media platforms, particularly Telegram, play a pivotal role in shaping public perceptions and opinions on global and national issues. Unlike traditional news media, Telegram allows for the proliferation of user-generated content with minimal oversight, making it a significant venue for the spread of controversial and misinformative content. During the COVID-19 pandemic, Telegram's popularity surged in Singapore, a country with one of the highest rates of social media use globally. We leverage Singapore-based Telegram data to analyze information flows within groups focused on COVID-19 and climate change. Using k-means clustering, we identified distinct user archetypes, including Skeptic, Engaged Advocate, Observer, and Analyst, each contributing uniquely to the discourse. We developed a model to classify users into these clusters (Precision: Climate change: 0.99; COVID-19: 0.95). By identifying these user archetypes and examining their contributions to information dissemination, we sought to uncover patterns to inform effective strategies for combating misinformation and enhancing public discourse on pressing global issues., Comment: Incomplete data and modification in data analysis

Large language models (LLMs) have achieved remarkable performance in language understanding and generation tasks by leveraging vast amounts of online texts. Unlike conventional models, LLMs can adapt to new domains through prompt engineering without the need for retraining, making them suitable for various business functions, such as strategic planning, project implementation, and data-driven decision-making. However, their limitations in terms of bias, contextual understanding, and sensitivity to prompts raise concerns about their readiness for real-world applications. This paper thoroughly examines the usefulness and readiness of LLMs for business processes. The limitations and capacities of LLMs are evaluated through experiments conducted on four accessible LLMs using real-world data. The findings have significant implications for organizations seeking to leverage generative AI and provide valuable insights into future research directions. To the best of our knowledge, this represents the first quantified study of LLMs applied to core business operations and challenges.

We build interpretable and lightweight transformer-like neural networks by unrolling iterative optimization algorithms that minimize graph smoothness priors -- the quadratic graph Laplacian regularizer (GLR) and the $\ell_1$-norm graph total variation (GTV) -- subject to an interpolation constraint. The crucial insight is that a normalized signal-dependent graph learning module amounts to a variant of the basic self-attention mechanism in conventional transformers. Unlike "black-box" transformers that require learning of large key, query and value matrices to compute scaled dot products as affinities and subsequent output embeddings, resulting in huge parameter sets, our unrolled networks employ shallow CNNs to learn low-dimensional features per node to establish pairwise Mahalanobis distances and construct sparse similarity graphs. At each layer, given a learned graph, the target interpolated signal is simply a low-pass filtered output derived from the minimization of an assumed graph smoothness prior, leading to a dramatic reduction in parameter count. Experiments for two image interpolation applications verify the restoration performance, parameter efficiency and robustness to covariate shift of our graph-based unrolled networks compared to conventional transformers.

We examine two types of binary betting markets, whose primary goal is for profit (such as sports gambling) or to gain information (such as prediction markets). We articulate the interplay between belief and price-setting to analyse both types of markets, and show that the goals of maximising bookmaker profit and eliciting information are fundamentally incompatible. A key insight is that profit hinges on the deviation between (the distribution of) bettor and true beliefs, and that heavier tails in bettor belief distribution imply higher profit. Our algorithmic contribution is to introduce online learning methods for price-setting. Traditionally bookmakers update their prices rather infrequently, we present two algorithms that guide price updates upon seeing each bet, assuming very little of bettor belief distributions. The online pricing algorithm achieves stochastic regret of $\mathcal{O}(\sqrt{T})$ against the worst local maximum, or $ \mathcal{O}(\sqrt{T \log T}) $ with high probability against the global maximum under fair odds. More broadly, the inherent trade-off between profit and information-seeking in binary betting may inspire new understandings of large-scale multi-agent behaviour., Comment: ICML 2024

Domain-specific languages (DSLs) are integral to various software workflows. Such languages offer domain-specific optimizations and abstractions that improve code readability and maintainability. However, leveraging these languages requires developers to rewrite existing code using the specific DSL's API. While large language models (LLMs) have shown some success in automatic code transpilation, none of them provide any functional correctness guarantees on the transpiled code. Another approach for automating this task is verified lifting, which relies on program synthesis to find programs in the target language that are functionally equivalent to the source language program. While several verified lifting tools have been developed for various application domains, they are specialized for specific source-target languages or require significant expertise in domain knowledge to make the search efficient. In this paper, leveraging recent advances in LLMs, we propose an LLM-based approach (LLMLift) to building verified lifting tools. We use the LLM's capabilities to reason about programs to translate a given program into its corresponding equivalent in the target language. Additionally, we use LLMs to generate proofs for functional equivalence. We develop lifting-based compilers for {\em four different} DSLs targeting different application domains. Our approach not only outperforms previous symbolic-based tools in both the number of benchmarks transpiled and transpilation time, but also requires significantly less effort to build.

LiteBIRD, the next-generation cosmic microwave background (CMB) experiment, aims for a launch in Japan's fiscal year 2032, marking a major advancement in the exploration of primordial cosmology and fundamental physics. Orbiting the Sun-Earth Lagrangian point L2, this JAXA-led strategic L-class mission will conduct a comprehensive mapping of the CMB polarization across the entire sky. During its 3-year mission, LiteBIRD will employ three telescopes within 15 unique frequency bands (ranging from 34 through 448 GHz), targeting a sensitivity of 2.2\,$\mu$K-arcmin and a resolution of 0.5$^\circ$ at 100\,GHz. Its primary goal is to measure the tensor-to-scalar ratio $r$ with an uncertainty $\delta r = 0.001$, including systematic errors and margin. If $r \geq 0.01$, LiteBIRD expects to achieve a $>5\sigma$ detection in the $\ell=$2-10 and $\ell=$11-200 ranges separately, providing crucial insight into the early Universe. We describe LiteBIRD's scientific objectives, the application of systems engineering to mission requirements, the anticipated scientific impact, and the operations and scanning strategies vital to minimizing systematic effects. We will also highlight LiteBIRD's synergies with concurrent CMB projects., Comment: 23 pages, 9 figures, 1 table, SPIE Astronomical Telescopes + Instrumentation 2024

We show that the Veneziano amplitude of string theory is the unique solution to an analytically solvable bootstrap problem. Uniqueness follows from two assumptions: faster than power-law falloff in high-energy scattering and the existence of some infinite sequence in momentum transfer at which higher-spin exchanges cancel. The string amplitude$\unicode{x2013}$including the mass spectrum$\unicode{x2013}$is an output of this bootstrap. If the amplitude merely vanishes at high energies, the solution is a three-parameter family containing the Veneziano, Coon, and hypergeometric amplitudes, and more., Comment: 5 pages (+ 2 pages supplementary), 1 figure

The adoption of Building Information Modeling (BIM) is beneficial in construction projects. However, it faces challenges due to the lack of a unified and scalable framework for converting 3D model details into BIM. This paper introduces SRBIM, a unified semantic reconstruction architecture for BIM generation. Our approach's effectiveness is demonstrated through extensive qualitative and quantitative evaluations, establishing a new paradigm for automated BIM modeling., Comment: CVPRW 2024, Oral

Precise segmentation of architectural structures provides detailed information about various building components, enhancing our understanding and interaction with our built environment. Nevertheless, existing outdoor 3D point cloud datasets have limited and detailed annotations on architectural exteriors due to privacy concerns and the expensive costs of data acquisition and annotation. To overcome this shortfall, this paper introduces a semantically-enriched, photo-realistic 3D architectural models dataset and benchmark for semantic segmentation. It features 4 different building purposes of real-world buildings as well as an open architectural landscape in Hong Kong. Each point cloud is annotated into one of 14 semantic classes., Comment: CVPRW 2024 (Oral)

Early-type galaxies (ETGs) are known to harbour dense spheroids of stars but scarce star formation (SF). Approximately a quarter of these galaxies have rich molecular gas reservoirs yet do not form stars efficiently. We study here the ETG NGC~524, with strong shear suspected to result in a smooth molecular gas disc and low star-formation efficiency (SFE). We present new spatially-resolved observations of the \textsuperscript{12}CO(2-1)-emitting cold molecular gas from the Atacama Large Millimeter/sub-millimeter Array (ALMA) and of the warm ionised-gas emission lines from SITELLE at the Canada-France-Hawaii Telescope. Although constrained by the resolution of the ALMA observations ($\approx37$~pc), we identify only $52$ GMCs with radii ranging from $30$ to $140$~pc, a low mean molecular gas mass surface density $\langle\Sigma_{\rm gas}\rangle\approx125$~M$_\odot$~pc$^{-2}$ and a high mean virial parameter $\langle\alpha_{\rm obs,vir}\rangle\approx5.3$. We measure spatially-resolved molecular gas depletion times ($\tau_{\rm dep}\equiv1/{\rm SFE}$) with a spatial resolution of $\approx100$~pc within a galactocentric distance of $1.5$~kpc. The global depletion time is $\approx2.0$~Gyr but $\tau_{\rm dep}$ increases toward the galaxy centre, with a maximum $\tau_{\rm dep,max}\approx5.2$~Gyr. However, no pure \ion{H}{II} region is identified in NGC~524 using ionised-gas emission-line ratio diagnostics, so the $\tau_{\rm dep}$ inferred are in fact lower limits. Measuring the GMC properties and dynamical states, we conclude that shear is the dominant mechanism shaping the molecular gas properties and regulating SF in NGC~524. This is supported by analogous analyses of the GMCs in a simulated ETG similar to NGC~524., Comment: 17 pages, 10 figures. To be published in MNRAS, accepted on May 27

In this paper, we propose a data-driven framework for collaborative wideband spectrum sensing and scheduling for networked unmanned aerial vehicles (UAVs), which act as the secondary users (SUs) to opportunistically utilize detected "spectrum holes". Our overall framework consists of three main stages. Firstly, in the model training stage, we explore dataset generation in a multi-cell environment and training a machine learning (ML) model using the federated learning (FL) architecture. Unlike the existing studies on FL for wireless that presume datasets are readily available for training, we propose a novel architecture that directly integrates wireless dataset generation, which involves capturing I/Q samples from over-the-air signals in a multi-cell environment, into the FL training process. Secondly, in the collaborative spectrum inference stage, we propose a collaborative spectrum fusion strategy that is compatible with the unmanned aircraft system traffic management (UTM) ecosystem. Finally, in the spectrum scheduling stage, we leverage reinforcement learning (RL) solutions to dynamically allocate the detected spectrum holes to the secondary users. To evaluate the proposed methods, we establish a comprehensive simulation framework that generates a near-realistic synthetic dataset using MATLAB LTE toolbox by incorporating base-station~(BS) locations in a chosen area of interest, performing ray-tracing, and emulating the primary users channel usage in terms of I/Q samples. This evaluation methodology provides a flexible framework to generate large spectrum datasets that could be used for developing ML/AI-based spectrum management solutions for aerial devices., Comment: This is a preprint version submitted to IEEE Transactions on Machine learning in Communications and Networking. arXiv admin note: text overlap with arXiv:2308.05036

Twisted bilayers of two-dimensional materials have emerged as a highly tunable platform for studying broken symmetry phases. While most interest has been focused on emergent states in systems whose constituent monolayers do not feature broken symmetry states, assembling monolayers that exhibit ordered states into twisted bilayers can also give rise to interesting phenomena. Here, we use large-scale first-principles density-functional theory calculations to study the atomic structure of twisted bilayer $\mathrm{{N}b{S}e_2}$ whose constituent monolayers feature a charge density wave. We find that different charge density wave states coexist in the ground state of the twisted bilayer: monolayer-like $3\times 3$ triangular and hexagonal charge density waves are observed in low-energy stacking regions, while stripe charge density waves are found in the domain walls surrounding the low-energy stacking regions. These predictions, which can be tested by scanning tunneling microscopy experiments, highlight the potential to create complex charge density wave ground states in twisted bilayer systems and can serve as a starting point for understanding superconductivity occurring at low temperatures.

We calculate the production and polarization of direct $J/\psi$ in the improved color evaporation model in $e+p$ photoproduction. We present the production as functions of transverse momentun, mass of the hadronic final state, and inelasticity. We also present the polarization parameters $\lambda_\vartheta$, $\lambda_{\varphi}$, and $\lambda_{\vartheta \varphi}$ in the helicity and the Collins-Soper frames, as well as the frame-invariant polarization parameter $\tilde{\lambda}$ as a function of transverse momentum and inelasticity. We find agreement with both $J/\psi$ unpolarized cross sections and the invariant polarization parameters as a function of $p_T$., Comment: arXiv admin note: substantial text overlap with arXiv:2102.09118

In cross-domain few-shot classification, \emph{nearest centroid classifier} (NCC) aims to learn representations to construct a metric space where few-shot classification can be performed by measuring the similarities between samples and the prototype of each class. An intuition behind NCC is that each sample is pulled closer to the class centroid it belongs to while pushed away from those of other classes. However, in this paper, we find that there exist high similarities between NCC-learned representations of two samples from different classes. In order to address this problem, we propose a bi-level optimization framework, \emph{maximizing optimized kernel dependence} (MOKD) to learn a set of class-specific representations that match the cluster structures indicated by labeled data of the given task. Specifically, MOKD first optimizes the kernel adopted in \emph{Hilbert-Schmidt independence criterion} (HSIC) to obtain the optimized kernel HSIC (opt-HSIC) that can capture the dependence more precisely. Then, an optimization problem regarding the opt-HSIC is addressed to simultaneously maximize the dependence between representations and labels and minimize the dependence among all samples. Extensive experiments on Meta-Dataset demonstrate that MOKD can not only achieve better generalization performance on unseen domains in most cases but also learn better data representation clusters. The project repository of MOKD is available at: \href{https://github.com/tmlr-group/MOKD}{https://github.com/tmlr-group/MOKD}.

Source-Free Domain Adaptation (SFDA) aims to adapt a pre-trained source model to a target domain using only unlabeled target data. Current SFDA methods face challenges in effectively leveraging pre-trained knowledge and exploiting target domain data. Multimodal Large Language Models (MLLMs) offer remarkable capabilities in understanding visual and textual information, but their applicability to SFDA poses challenges such as instruction-following failures, intensive computational demands, and difficulties in performance measurement prior to adaptation. To alleviate these issues, we propose Reliability-based Curriculum Learning (RCL), a novel framework that integrates multiple MLLMs for knowledge exploitation via pseudo-labeling in SFDA. Our framework incorporates proposed Reliable Knowledge Transfer, Self-correcting and MLLM-guided Knowledge Expansion, and Multi-hot Masking Refinement to progressively exploit unlabeled data in the target domain. RCL achieves state-of-the-art (SOTA) performance on multiple SFDA benchmarks, e.g., $\textbf{+9.4%}$ on DomainNet, demonstrating its effectiveness in enhancing adaptability and robustness without requiring access to source data. Code: https://github.com/Dong-Jie-Chen/RCL.

Instructing large language models (LLMs) to solve elementary school math problems has shown great success using Chain of Thought (CoT). However, the CoT approach relies on an LLM to generate a sequence of arithmetic calculations which can be prone to cascaded calculation errors. We hypothesize that an LLM should focus on extracting predicates and generating symbolic formulas from the math problem description so that the underlying calculation can be done via an external code interpreter. We investigate using LLM to generate Prolog programs to solve mathematical questions. Experimental results show that our Prolog-based arithmetic problem-solving outperforms CoT generation in the GSM8K benchmark across three distinct LLMs. In addition, given the insensitive ordering of predicates and symbolic formulas in Prolog, we propose to permute the ground truth predicates for more robust LLM training via data augmentation., Comment: 12 pages, 4 figures, accepted by NAACL 2024 Main Conference

As the importance of comprehensive evaluation in workshop courses increases, there is a growing demand for efficient and fair assessment methods that reduce the workload for faculty members. This paper presents an evaluation conducted with Large Language Models (LLMs) using actual student essays in three scenarios: 1) without providing guidance such as rubrics, 2) with pre-specified rubrics, and 3) through pairwise comparison of essays. Quantitative analysis of the results revealed a strong correlation between LLM and faculty member assessments in the pairwise comparison scenario with pre-specified rubrics, although concerns about the quality and stability of evaluations remained. Therefore, we conducted a qualitative analysis of LLM assessment comments, showing that: 1) LLMs can match the assessment capabilities of faculty members, 2) variations in LLM assessments should be interpreted as diversity rather than confusion, and 3) assessments by humans and LLMs can differ and complement each other. In conclusion, this paper suggests that LLMs should not be seen merely as assistants to faculty members but as partners in evaluation committees and outlines directions for further research.

Accurate surface roughness prediction is critical for ensuring high product quality, especially in areas like manufacturing and aerospace, where the smallest imperfections can compromise performance or safety. However, this is challenging due to complex, non-linear interactions among variables, which is further exacerbated with limited and imbalanced datasets. Existing methods using traditional machine learning algorithms require extensive domain knowledge for feature engineering and substantial human intervention for model selection. To address these issues, we propose NASPrecision, a Neural Architecture Search (NAS)-Driven Multi-Stage Learning Framework. This innovative approach autonomously identifies the most suitable features and models for various surface roughness prediction tasks and significantly enhances the performance by multi-stage learning. Our framework operates in three stages: 1) architecture search stage, employing NAS to automatically identify the most effective model architecture; 2) initial training stage, where we train the neural network for initial predictions; 3) refinement stage, where a subsequent model is appended to refine and capture subtle variations overlooked by the initial training stage. In light of limited and imbalanced datasets, we adopt a generative data augmentation technique to balance and generate new data by learning the underlying data distribution. We conducted experiments on three distinct real-world datasets linked to different machining techniques. Results show improvements in Mean Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE), and Standard Deviation (STD) by 18%, 31%, and 22%, respectively. This establishes it as a robust and general solution for precise surface roughness prediction, potentially boosting production efficiency and product quality in key industries while minimizing domain expertise and human intervention.

We introduce and analyze a model for self-reconfigurable robots made up of unit-cube modules. Compared to past models, our model aims to newly capture two important practical aspects of real-world robots. First, modules often do not occupy an exact unit cube, but rather have features like bumps extending outside the allotted space so that modules can interlock. Thus, for example, our model forbids modules from squeezing in between two other modules that are one unit distance apart. Second, our model captures the practical scenario of many passive modules assembled by a single robot, instead of requiring all modules to be able to move on their own. We prove two universality results. First, with a supply of auxiliary modules, we show that any connected polycube structure can be constructed by a carefully aligned plane sweep. Second, without additional modules, we show how to construct any structure for which a natural notion of external feature size is at least a constant; this property largely consolidates forbidden-pattern properties used in previous works on reconfigurable modular robots.

We argue that representations in AI models, particularly deep networks, are converging. First, we survey many examples of convergence in the literature: over time and across multiple domains, the ways by which different neural networks represent data are becoming more aligned. Next, we demonstrate convergence across data modalities: as vision models and language models get larger, they measure distance between datapoints in a more and more alike way. We hypothesize that this convergence is driving toward a shared statistical model of reality, akin to Plato's concept of an ideal reality. We term such a representation the platonic representation and discuss several possible selective pressures toward it. Finally, we discuss the implications of these trends, their limitations, and counterexamples to our analysis., Comment: Equal contributions. Project: https://phillipi.github.io/prh/ Code: https://github.com/minyoungg/platonic-rep

Model Inversion (MI) attacks aim to reconstruct private training data by abusing access to machine learning models. Contemporary MI attacks have achieved impressive attack performance, posing serious threats to privacy. Meanwhile, all existing MI defense methods rely on regularization that is in direct conflict with the training objective, resulting in noticeable degradation in model utility. In this work, we take a different perspective, and propose a novel and simple Transfer Learning-based Defense against Model Inversion (TL-DMI) to render MI-robust models. Particularly, by leveraging TL, we limit the number of layers encoding sensitive information from private training dataset, thereby degrading the performance of MI attack. We conduct an analysis using Fisher Information to justify our method. Our defense is remarkably simple to implement. Without bells and whistles, we show in extensive experiments that TL-DMI achieves state-of-the-art (SOTA) MI robustness. Our code, pre-trained models, demo and inverted data are available at: https://hosytuyen.github.io/projects/TL-DMI