arXiv:2605.27310v1 Announce Type: new
Abstract: Cross-view spatial reasoning remains a weak spot for vision-language models (VLMs): they often reason in language and lose the fine-grained geometry needed for the task. Thinking with images aims to address this by generating an intermediate thinking image, but recent work shows that models often ignore the visual evidence in these traces. We therefore ask how to make visual thinking matter, and what kind of visual thinking works best. We study these questions in unified multimodal models (UMMs), which natively support interleaved image-text generation. For the first question, we propose View Dropout (VDrop), a training-time intervention that hides parts of one input view from the answer span while keeping them visible to the thinking-image tokens. This encourages the model to use the thinking image when answering, instead of relying only on the input views. Once the thinking image is used for answer prediction, we study which type of visual thinking is most effective. We frame this as a learnability-informativeness tradeoff and compare three thinking-image variants: top-down, panoramic, and point-matching renderings. Trained on synthetic scenes and evaluated on five real-world out-of-domain benchmarks, panoramic visual thinking with VDrop is the only configuration that is both informative and learnable, and it achieves the best out-of-domain generalization.
Science Journals
arXiv:2605.27311v1 Announce Type: new
Abstract: Chart question-answering (QA) benchmarks aim to pose questions that require visual reasoning to correctly answer, but models can often reach solutions through shortcuts or prior familiarity with a chart based on their own background knowledge. To strictly evaluate visual reasoning, we propose counterfactual charts where the chart-question task remains fixed, but underlying chart and the corresponding answer are varied. We introduce Chartographer, a framework to reverse engineer charts into executable code, validate reconstruction fidelity, generate seed-controlled counterfactual variants, and derive new answers from executable QA logic. We apply this framework to existing chart QA datasets and evaluate proprietary and open-source vision-language models (VLMs), measuring variation sensitivity and generalizability. Counterfactual charts reveal failures hidden by single-chart performance: VLMs often fail to generalize after answering the original chart correctly. We find failures are most prevalent when updated charts require novel visual reasoning pathways.
arXiv:2605.27312v1 Announce Type: new
Abstract: The automorphism group of a code is the group of permutations that map a code to itself. Berman codes are a class of binary linear codes characterized by two integer parameters $n\geq 2$ and $m\geq 1$, and this class includes the Reed-Muller codes as well. The class of Berman codes and their duals were recently shown to achieve the capacity of the binary erasure channel. A number of abelian codes that arise from the intersection and subspace sums of Berman and Dual Berman codes were also identified recently, for odd $n\geq 3$. A subclass of these abelian codes was shown to have good short block-length performance for AWGN channels, with efficient decoding algorithms. In this work, we identify the exact automorphism group for Berman codes and their duals. Further, we find the exact automorphism group for the above mentioned abelian codes, when $n\geq 5$. In the case of such abelian codes with $n=3$, we present partial characterizations of the automorphism groups for a large collection of parameter choices, and complete characterizations for a few.
arXiv:2605.27314v1 Announce Type: new
Abstract: Reactive control is often considered insufficient for multi-objective tasks because conflicting objectives give rise to local minima. We argue this limitation is not inherent but arises from static encodings that fail to reflect how objectives currently interact. We exploit the interaction structure encoded in a graph-based world model by extending it with nullspace projections: conflicts are resolved where they arise by projecting lower-priority gradients into the nullspace of higher-priority ones, with priorities determined continuously from the current state. We demonstrate this in two domains where conflicts between objectives are central: navigation around non-convex obstacles, where static potential fields fundamentally fail, and planar pushing of non-convex objects, where our method achieves $100\%$ success across one-hundred configurations versus $0\%$ for the steepest-descent baseline and ${\sim}55\%$ for diffusion policy, without demonstrations or retraining. The same formulation transfers directly to a real robot with additional perceptual and kinematic constraints, accommodating them through the same mechanism.
arXiv:2605.27205v1 Announce Type: cross
Abstract: Wireless digital twins require repeated synchronization between a time-evolving physical scene and its digital counterpart under limited and time-varying communication resources. For perception-centric twins, pixel-domain transmission or uniformly protected bitstreams can be mismatched to the semantic state consumed by twin-side applications. This paper proposes TWIST, a closed-loop token synchronization framework for application-aware wireless digital twins. TWIST represents each physical observation as a token and synchronizes this state over a wireless link, rather than optimizing visual reconstruction. Token positions are grouped by task relevance and protected through mode-conditioned unequal error protection under low-, medium-, and high-synchronization modes. At the twin side, decoding confidence converts unreliable hard token decisions into erasures, which are restored by a completion model before updating the semantic twin state. The recovered state supports traffic-state inference and generates compact feedback statistics, including channel quality, receiver uncertainty, semantic drift, and application priority, for subsequent mode adaptation. Experiments on a dynamic road-scene digital-twin scenario show that TWIST improves traffic-state inference and semantic twin-state synchronization compared with fixed-mode and channel-only adaptation strategies, while reducing the average synchronization cost relative to always-high transmission.
arXiv:2605.27231v1 Announce Type: cross
Abstract: Scalar Induced Gravitational Waves (SIGW) are generated at second order in perturbation theory and to achieve observational relevance, inflationary dynamics must evade the standard slow-roll scenario at small scales, generating large curvature perturbations following strongly non-Gaussian statistics. We propose a method to efficiently compute the SIGW spectrum including arbitrary non-Gaussianities. First, we solve the wave equation adopting semi-analytic methods; this results in an expression involving integrals in Fourier space which are impossible to solve directly on a lattice. We overcome this bottleneck by recasting these integrals as a sum of about 50 convolutions, each of which can be computed efficiently with FFT methods. Finally, the power spectrum is measured directly from the lattice realization. We implement this in FLAN-SIGW, a GPU-accelerated code capable of computing fully non-perturbative, non-Gaussian SIGW spectra in seconds with an error within 10% with modest computational resources. The code is made public at https://github.com/giovannipiccoli99/FLAN-SIGW. In this first implementation, in order to assess the performance of the method, we adopt a standard radiation-dominated background with $w = 1/3$.
arXiv:1705.03307v5 Announce Type: replace
Abstract: We define and develop two-level type theory (2LTT), a version of Martin-L\"of type theory which combines two different type theories. We refer to them as the inner and the outer type theory. In our case of interest, the inner theory is homotopy type theory (HoTT) which may include univalent universes and higher inductive types. The outer theory is a traditional form of type theory validating uniqueness of identity proofs (UIP). One point of view on it is as internalised meta-theory of the inner type theory.
There are two motivations for 2LTT. Firstly, there are certain results about HoTT which are of meta-theoretic nature, such as the statement that semisimplicial types up to level $n$ can be constructed in HoTT for any externally fixed natural number $n$. Such results cannot be expressed in HoTT itself, but they can be formalised and proved in 2LTT, where $n$ will be a variable in the outer theory. This point of view is inspired by observations about conservativity of presheaf models.
Secondly, 2LTT is a framework which is suitable for formulating additional axioms that one might want to add to HoTT. This idea is heavily inspired by Voevodsky's Homotopy Type System (HTS), which constitutes one specific instance of a 2LTT. HTS has an axiom ensuring that the type of natural numbers behaves like the external natural numbers, which allows the construction of a universe of semisimplicial types. In 2LTT, this axiom can be stated simply be asking the inner and outer natural numbers to be isomorphic.
After defining 2LTT, we set up a collection of tools with the goal of making 2LTT a convenient language for future developments. As a first such application, we develop the theory of Reedy fibrant diagrams in the style of Shulman. Continuing this line of thought, we suggest a definition of (infinity,1)-category and give some examples.
arXiv:2210.02573v5 Announce Type: replace
Abstract: Learning the physical simulation on large-scale meshes with flat Graph Neural Networks (GNNs) and stacking Message Passings (MPs) is challenging due to the scaling complexity w.r.t. the number of nodes and over-smoothing. There has been growing interest in the community to introduce \textit{multi-scale} structures to GNNs for physical simulation. However, current state-of-the-art methods are limited by their reliance on the labor-intensive drawing of coarser meshes or building coarser levels based on spatial proximity, which can introduce wrong edges across geometry boundaries. Inspired by the bipartite graph determination, we propose a novel pooling strategy, \textit{bi-stride} to tackle the aforementioned limitations. Bi-stride pools nodes on every other frontier of the breadth-first search (BFS), without the need for the manual drawing of coarser meshes and avoiding the wrong edges by spatial proximity. Additionally, it enables a one-MP scheme per level and non-parametrized pooling and unpooling by interpolations, resembling U-Nets, which significantly reduces computational costs. Experiments show that the proposed framework, \textit{BSMS-GNN}, significantly outperforms existing methods in terms of both accuracy and computational efficiency in representative physical simulations.
arXiv:2302.13473v2 Announce Type: replace
Abstract: Federated learning (FL) enables multiple data owners to build machine learning models collaboratively without exposing their private local data. In order for FL to achieve widespread adoption, it is important to balance the need for performance, privacy-preservation and interpretability, especially in mission critical applications such as finance and healthcare. Thus, interpretable federated learning (IFL) has become an emerging topic of research attracting significant interest from the academia and the industry alike. Its interdisciplinary nature can be challenging for new researchers to pick up. In this paper, we bridge this gap by providing (to the best of our knowledge) the first survey on IFL. We propose a unique IFL taxonomy which covers relevant works enabling FL models to explain the prediction results, support model debugging, and provide insights into the contributions made by individual data owners or data samples, which in turn, is crucial for allocating rewards fairly to motivate active and reliable participation in FL. We conduct comprehensive analysis of the representative IFL approaches, the commonly adopted performance evaluation metrics, and promising directions towards building versatile IFL techniques.
arXiv:2401.07669v3 Announce Type: replace
Abstract: Adapting CLIP for videos has gained popularity due to its semantic and rich representation. While CLIP is a good starting point, it typically undergoes post-pretraining (contrastive finetuning) on large video narration or caption datasets (e.g. HowTo100M, WebVid2.5M). However, such narrations or captions often lack comprehensive information needed to represent a video holistically. As the learning signal from text is sparse, the visual learning is inefficient and adaptation requires millions of samples to post-pretrain. In this work, we ask: is it possible to efficiently adapt CLIP for general and holistic video understanding? We use videos labeled with structured and dense Semantic Role Labels (SRLs) that capture actions, people or objects, their attributes, adverbs (manner), and location in a structured format representing the entire video in a holistic way. We generate rule-based captions from SRLs and demonstrate that simple contrastive finetuning on a mere 23k video-caption pairs is adequate to learn powerful, transferable representations applicable across a diverse range of video understanding tasks that require varying levels of perceptual granularity. Our adapted CLIP model, SRL-CLIP, exhibits comparable or superior performance on zero-shot text-to-video retrieval compared to state-of-the-art models that possess 4-8x more parameters and are post-pretrained on up to 6000x more data. SRL-CLIP surpasses CLIP on multiple video benchmarks, underscoring the efficient learning and improved representations.
arXiv:2403.04976v4 Announce Type: replace
Abstract: Recent improvements in energy efficiency and renewable energy integration have increased the relative importance of embodied carbon in data centers, motivating improved provisioning strategies. Conventional approaches primarily minimize operational energy, but this perspective is increasingly insufficient for sustainability. In this paper, we propose carbon depreciation models to encourage longer hardware lifetimes. Carbon depreciation assigns a larger portion of embodied carbon to newly provisioned servers, discouraging unnecessary deployment of new hardware. As a result, new servers are provisioned mainly for jobs with strict quality-of-service (QoS) constraints, while older servers, whose embodied carbon has largely been recovered, are used for other workloads. We further argue that both embodied carbon and operational carbon from server idle time should be recovered during active jobs, encouraging provisioning strategies that maintain high utilization. We show that prior carbon accounting strategies can be counterproductive: under a greedy scheduler minimizing carbon under QoS constraints, jobs are priced as 25% cheaper on new hardware than on older hardware. In contrast, our approach uses a greedy scheduler that prioritizes older hardware through non-linear carbon depreciation, promoting sustainable provisioning. Experimental results show carbon reductions of 28-57%, depending on server lifetime assumptions.
arXiv:2405.16417v2 Announce Type: replace
Abstract: Recent vision-language pre-trained models (VL-PTMs) have shown remarkable success in open-vocabulary tasks. However, downstream use cases often involve further fine-tuning of VL-PTMs, which may distort their general knowledge and impair their ability to handle distribution shifts. In real-world scenarios, machine learning systems inevitably encounter both covariate shifts (e.g., changes in image styles) and semantic shifts (e.g., test-time unseen classes). This highlights the importance of enhancing out-of-distribution (OOD) generalization on covariate shifts and simultaneously detecting semantic-shifted unseen classes. Thus a critical but underexplored question arises: How to improve VL-PTMs' generalization ability to closed-set OOD data, while effectively detecting open-set unseen classes during fine-tuning? In this paper, we propose a novel objective function of OOD detection that also serves to improve OOD generalization. We show that minimizing the gradient magnitude of energy scores on training data leads to domain-consistent Hessians of classification loss, a strong indicator for OOD generalization revealed by theoretical analysis. Based on this finding, we have developed a unified fine-tuning framework that allows for concurrent optimization of both tasks. Extensive experiments have demonstrated the superiority of our method. The code is available at https://github.com/LinLLLL/CRoFT.
arXiv:2406.03474v2 Announce Type: replace
Abstract: Language-guided autonomous driving requires bridging a large abstraction gap between high-level natural-language instructions and low-level vehicle control. End-to-end approaches that use a single multimodal large language model (MLLM) to map language directly to actions struggle with this mismatch, often failing to exploit the reasoning capabilities of the model and exhibiting limited generalization beyond the distributions of driving datasets used for fine-tuning. To address this issue, we propose AD-H, a hierarchical multi-agent framework that explicitly separates high-level decision-making from low-level vehicle execution. At the upper level, an MLLM-based planner interprets natural-language commands and environmental context to generate coherent mid-level driving instructions. At the lower level, a lightweight controller converts these mid-level instructions into precise, continuous control actions. This decomposition aligns with the functional strengths of each component: the planner focuses on semantic reasoning and task decomposition, while the controller ensures stable and accurate actuation. To support large-scale training under this hierarchy, we design a rule-based pipeline that reconstructs mid-level commands from driving signals, producing 1.15 million hierarchical annotation pairs. Extensive experiments show that AD-H outperforms state-of-the-art models despite using fewer parameters, namely 3B plus 350M compared with 7B, and achieves superior long-horizon generalization and instruction-following performance. We make our data and code publicly accessible at https://github.com/zhangzaibin/AD-H
arXiv:2406.20056v2 Announce Type: replace
Abstract: We extend the notion of activity for automaton semigroups and monoids introduced by Bartholdi, Godin, Klimann and Picantin to a more general setting. Their activity notion was already a generalization of Sidki's activity hierarchy for automaton groups. We show that the language of $\omega$-words with infinite orbits is effectively a deterministic B\"uchi language for automata with bounded extended activity, which yields decidability of the finiteness problem for complete automaton semigroups and monoids of bounded activity (solving an open problem by Bartholdi, Godin, Klimann and Picantin). In fact, we obtain a stronger result also covering finitely generated subsemigroups.
arXiv:2407.15073v4 Announce Type: replace
Abstract: Causal discovery aims to identify causal relationships between variables and is a fundamental problem across the sciences. Traditional statistical causal discovery (SCD) methods rely solely on observational data and ignore the contextual information available in metadata, whereas recent LLM-based methods exploit metadata but treat the large language model (LLM) as a single agent, leaving its judgments vulnerable to memorized or biased associations. To address this gap, we introduce MAC (Multi-Agent Causal Discovery Framework), which casts causal discovery as a multi-agent debate coupled with the autonomous selection of an SCD algorithm. MAC combines two complementary modules, bridged by a Meta Fusion mechanism: a Debate-Coding Module (DCM) that grounds an initial graph in data by autonomously selecting and executing the best-suited SCD algorithm, and a Meta-Debate Module (MDM) that refines the graph through an adversarial Affirmative-Negative-Judge debate over the metadata. Across five benchmark datasets and three metrics (F1, SHD, NHD), MAC achieves the best aggregate performance among five statistical and four LLM-based baselines, ranking first on 10 of 15 evaluation points with Gemini-2.0-Flash -- including a perfect reconstruction of the Earthquake graph -- and remains robust across three backbone LLMs.
arXiv:2408.08946v3 Announce Type: replace
Abstract: Accurate attribution of authorship is crucial for maintaining the integrity of digital content, improving forensic investigations, and mitigating the risks of misinformation and plagiarism. Addressing the imperative need for proper authorship attribution is essential to uphold the credibility and accountability of authentic authorship. The rapid advancements of Large Language Models (LLMs) have blurred the lines between human and machine authorship, posing significant challenges for traditional methods. We present a comprehensive literature review that examines the latest research on authorship attribution in the era of LLMs. This survey systematically explores the landscape of this field by categorizing four representative problems: (1) Human-written Text Attribution; (2) LLM-generated Text Detection; (3) LLM-generated Text Attribution; and (4) Human-LLM Co-authored Text Attribution. We also discuss the challenges related to ensuring the generalization and explainability of authorship attribution methods. Generalization requires the ability to generalize across various domains, while explainability emphasizes providing transparent and understandable insights into the decisions made by these models. By evaluating the strengths and limitations of existing methods and benchmarks, we identify key open problems and future research directions in this field. This literature review serves as a roadmap for researchers and practitioners interested in understanding the state of the art in this rapidly evolving field.
arXiv:2408.15787v2 Announce Type: replace
Abstract: Creating effective dialogue systems for mental health support requires high-quality multi-turn counseling dialogue data, yet collecting real counselor-client conversations presents significant challenges, including privacy concerns, high costs, and limited scalability. We present \textbf{Interactive Agents}, a novel framework that simulates naturalistic counseling dialogues through controlled LLM-to-LLM interactions. The framework introduces two key innovations: (1) a personalized client agent that maintains consistent psychological characteristics throughout a session, and (2) a counselor agent that implements a theoretically grounded three-stage therapeutic model comprising the exploration, insight, and action phases. Through rigorous evaluation using both automatic metrics and professional-counselor assessments based on the Working Alliance Inventory, we demonstrate that our framework generates therapeutically valid dialogues that are comparable in quality to human-generated sessions. Models fine-tuned on our proposed synthetic dataset (SimPsyDial) achieve state-of-the-art performance in a standard pairwise chatbot-arena evaluation of LLM-based counselors. Our framework provides a scalable, privacy-preserving method for generating high-quality counseling dialogue data while maintaining professional therapeutic standards.
arXiv:2410.00357v2 Announce Type: replace
Abstract: Neural scaling laws play a pivotal role in the performance of deep neural networks and have been observed in a wide range of tasks. However, a complete theoretical framework for understanding these scaling laws remains underdeveloped. In this paper, we explore the neural scaling laws for deep operator networks, which involve learning mappings between function spaces, with a focus on the Chen and Chen style architecture. These approaches, which include the popular Deep Operator Network (DeepONet), approximate the output functions using a linear combination of learnable basis functions and coefficients that depend on the input functions. We establish a theoretical framework to quantify the neural scaling laws by analyzing its approximation and generalization errors. We articulate the relationship between the approximation and generalization errors of deep operator networks and key factors such as network model size and training data size. Moreover, we address cases where input functions exhibit low-dimensional structures, allowing us to derive tighter error bounds. These results also hold for deep ReLU networks and other similar structures. Our results offer a partial explanation of the neural scaling laws in operator learning and provide a theoretical foundation for their applications.
arXiv:2410.00995v3 Announce Type: replace
Abstract: The automatic synthesis of analog circuits presents significant challenges. Most existing approaches formulate the problem as a single-objective optimization task, overlooking that design specifications for a given circuit type vary widely across applications. To address this, we introduce specification-conditioned analog circuit generation, a task that directly generates analog circuits based on target specifications. The motivation is to leverage existing well-designed circuits to improve automation in analog circuit design. Specifically, we propose CktGen, a simple yet effective variational autoencoder that maps discretized specifications and circuits into a joint latent space and reconstructs the circuit from that latent vector. Notably, as a single specification may correspond to multiple valid circuits, naively fusing specification information into the generative model does not capture these one-to-many relationships. To address this, we decouple the encoding of circuits and specifications and align their mapped latent space. Then, we employ contrastive training with a filter mask to maximize differences between encoded circuits and specifications. Furthermore, classifier guidance along with latent feature alignment promotes the clustering of circuits sharing the same specification, avoiding model collapse into trivial one-to-one mappings. By canonicalizing the latent space with respect to specifications, we can search for an optimal circuit that meets valid target specifications. We conduct comprehensive experiments on the open circuit benchmark and introduce metrics to evaluate cross-model consistency. Experimental results demonstrate that CktGen achieves substantial improvements over state-of-the-art methods.
arXiv:2410.21536v4 Announce Type: replace
Abstract: In this paper, we propose Intelligent Environmental Empathy (IEE) as a new driver for climate peace and justice, as an emerging issue in the age of big data. We first show that the authoritarian top-down intergovernmental cooperation, through international organizations (e.g., UNEP) for climate justice, could not overcome environmental issues and crevices so far. We elaborate on four grounds of climate injustice (i.e., teleological origin, axiological origin, formation cause, and social epistemic cause), and explain how the lack of empathy and environmental motivation on a global scale causes the failure of all the authoritarian top-down intergovernmental cooperation. Addressing all these issues requires a new button-up approach to climate peace and justice. Secondly, focusing on the intersection of AI, environmental empathy, and climate justice, we propose a model of Intelligent Environmental Empathy (IEE) for climate peace and justice at the operational level. IEE is empowered by the new power of environmental empathy (as a driver of green obligation for climate justice) and putative decentralized platform of AI (as an operative system against free riders), which Initially, impact citizens and some middle-class decision makers, such as city planners and local administrators, but will eventually affect global decision-makers as well.
arXiv:2411.02355v4 Announce Type: replace
Abstract: Quantization is a powerful tool for accelerating large language model (LLM) inference, but the accuracy-performance trade-offs across different formats remain unclear. In this paper, we conduct the most comprehensive empirical study to date, evaluating FP8, INT8, and INT4 quantization across academic benchmarks and real-world tasks on the entire Llama-3.1 model family. Through over 500,000 evaluations, our investigation yields several key findings: (1) FP8 (W8A8-FP) is effectively lossless across all model scales, (2) well-tuned INT8 (W8A8-INT) achieves surprisingly low (1-3\%) accuracy degradation, and (3) INT4 weight-only (W4A16-INT) is more competitive than expected, rivaling 8-bit quantization. Further, we investigate the optimal quantization format for different deployments by analyzing inference performance through the popular vLLM framework. Our analysis provides clear deployment recommendations: W4A16 is the most cost-efficient for synchronous setups, while W8A8 dominates in asynchronous continuous batching. For mixed workloads, the optimal choice depends on the specific use case. Our findings offer practical, data-driven guidelines for deploying quantized LLMs at scale -- ensuring the best balance between speed, efficiency, and accuracy.
arXiv:2412.20505v2 Announce Type: replace
Abstract: Participatory Urban Planning (PUP) is increasingly supported by LLM-based agents, yet existing methods largely rely on static preference elicitation and one-shot stakeholder discussions, overlooking the cyclical nature of real-world planning, where residential life, experience collection, and plan adjustment continually interact. We propose Living-in-the-loop Participatory Urban Planning (LiPUP), a closed-loop paradigm that alternates between simulated residential living and experience-driven plan revision, while posing two key challenges: grounding scattered living experience in concrete urban contexts and translating subjective feedback into spatially coherent planning actions. To instantiate LiPUP, we introduce LiPUP-MA, an LLM-based multi-agent framework that constructs a Plan-centric Graph-based Experience Bank to organize urban-grounded residential feedback from living simulation and equips a Spatially-constrained Skill-augmented Planner agent to revise plans by harmonizing experiential, visual, and geospatial evidence. Experiments show that LiPUP-MA consistently outperforms baselines on both conventional static planning metrics and living-based metrics, while iterative LiPUP cycles further improve plan quality.
arXiv:2501.00520v2 Announce Type: replace
Abstract: This paper presents a comprehensive study on the classification and detection of Silicosis-related lung inflammation. Our main contributions include 1) the creation of a newly curated chest X-ray (CXR) image dataset named SVBCX that is tailored to the nuances of lung inflammation caused by distinct agents, providing a valuable resource for silicosis and pneumonia research community; and 2) we propose a novel deep-learning architecture that integrates graph transformer networks alongside a traditional deep neural network module for the effective classification of silicosis and pneumonia. Additionally, we employ the Balanced Cross-Entropy (BalCE) as a loss function to ensure more uniform learning across different classes, enhancing the model's ability to discern subtle differences in lung conditions. The proposed model architecture and loss function selection aim to improve the accuracy and reliability of inflammation detection, particularly in the context of Silicosis. Furthermore, our research explores the efficacy of an ensemble approach that combines the strengths of diverse model architectures. Experimental results on the constructed dataset demonstrate promising outcomes, showcasing substantial enhancements compared to baseline models. The ensemble of models achieves a macro-F1 score of 0.9749 and AUC ROC scores exceeding 0.99 for each class, underscoring the effectiveness of our approach in accurate and robust lung inflammation classification.
arXiv:2501.06708v5 Announce Type: replace
Abstract: Large-scale web-crawled datasets contain noise, bias, and irrelevant information, necessitating data selection techniques. Existing methods depend on hand-crafted heuristics, downstream datasets, or require expensive influence-based computations -- all of which limit scalability and introduce unwanted data dependencies. To address this, we introduce the Mimic Score, a simple and geometry-based data-quality metric that evaluates utility by measuring alignment between a sample's gradients and a target direction induced by a pre-trained reference model. This leverages readily available model weights, avoids needing validation datasets, and incurs minimal computational overheads. Building on this metric, we propose Grad-Mimic, a two-stage framework that re-weights samples online to accelerate training and aggregates sample utilities offline to construct effective data filters. Empirically, we show that using mimic scores to guide training improves data efficiency, accelerates convergence, yields consistent performance gains across six image datasets, and enhances CLIP models with 20.7% fewer training steps. Additionally, mimic score-based filters augment existing filtering techniques, enabling improved CLIP models trained with 4.7 million fewer samples.
arXiv:2502.03946v5 Announce Type: replace
Abstract: Data preprocessing is often paid little attention in machine learning, despite its potentially significant impact on model performance. While automated machine learning pipelines are starting to recognize and integrate data preprocessing into their solutions for classification and regression tasks, this integration is lacking for more specialized tasks like time-to-event models for censored data. As a result, survival analysis not only faces the general challenges of data preprocessing but also suffers from the lack of tailored, automated solutions in this area. To address this gap, this paper presents CleanSurvival, a reinforcement-learning-based solution for optimizing preprocessing pipelines, extended specifically for survival analysis. The framework can handle continuous and categorical variables. It builds upon Learn2Clean's Q-learning to select which combination of data imputation, outlier detection and feature extraction techniques achieves optimal performance for a Cox, random forest, neural network or user-supplied time-to-event model. The Python package is available on GitHub: https://github.com/datasciapps/CleanSurvival. Experimental benchmarks on real-world datasets show that the Q-learning-based data preprocessing can improve predictive performance relative to simple baselines, while runtime behavior is condition-dependent and most clearly interpretable in the best-covered benchmark cells. Furthermore, a simulation study demonstrates effectiveness across different types and levels of missingness and noise. With an increase in the use of machine learning, it becomes important to generalise AutoML pipelines to a variety of models now present, including survival analysis. Tools like CleanSurvival, which integrate preprocessing for survival analysis, can make survival studies easier and quicker to perform, as well as make the results more robust.