Search

Adaptively forecasting human behavior in social settings is an important step toward achieving Artificial General Intelligence. Most existing research in social forecasting has focused either on unfocused interactions, such as pedestrian trajectory prediction, or on monadic and dyadic behavior forecasting. In contrast, social psychology emphasizes the importance of group interactions for understanding complex social dynamics. This creates a gap that we address in this paper: forecasting social interactions at the group (conversation) level. Additionally, it is important for a forecasting model to be able to adapt to groups unseen at train time, as even the same individual behaves differently across different groups. This highlights the need for a forecasting model to explicitly account for each group's unique dynamics. To achieve this, we adopt a meta-learning approach to human behavior forecasting, treating every group as a separate meta-learning task. As a result, our method conditions its predictions on the specific behaviors within the group, leading to generalization to unseen groups. Specifically, we introduce Social Process (SP) models, which predict a distribution over future multimodal cues jointly for all group members based on their preceding low-level multimodal cues, while incorporating other past sequences of the same group's interactions. In this work we also analyze the generalization capabilities of SP models in both their outputs and latent spaces through the use of realistic synthetic datasets., Comment: This is an extension paper to "Social Processes: Self-Supervised Meta-Learning over Conversational Groups for Forecasting Nonverbal Social Cues", by Raman et al. (arXiv:2107.13576)

While 2D pose estimation has advanced our ability to interpret body movements in animals and primates, it is limited by the lack of depth information, constraining its application range. 3D pose estimation provides a more comprehensive solution by incorporating spatial depth, yet creating extensive 3D pose datasets for animals is challenging due to their dynamic and unpredictable behaviours in natural settings. To address this, we propose a hybrid approach that utilizes rigged avatars and the pipeline to generate synthetic datasets to acquire the necessary 3D annotations for training. Our method introduces a simple attention-based MLP network for converting 2D poses to 3D, designed to be independent of the input image to ensure scalability for poses in natural environments. Additionally, we identify that existing anatomical keypoint detectors are insufficient for accurate pose retargeting onto arbitrary avatars. To overcome this, we present a lookup table based on a deep pose estimation method using a synthetic collection of diverse actions rigged avatars perform. Our experiments demonstrate the effectiveness and efficiency of this lookup table-based retargeting approach. Overall, we propose a comprehensive framework with systematically synthesized datasets for lifting poses from 2D to 3D and then utilize this to re-target motion from wild settings onto arbitrary avatars., Comment: 2025 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025)

Large language models (LLMs) are empowering decision-making in several applications, including tool or API usage and answering multiple-choice questions (MCQs). However, they often make overconfident, incorrect predictions, which can be risky in high-stakes settings like healthcare and finance. To mitigate these risks, recent works have used conformal prediction (CP), a model-agnostic framework for distribution-free uncertainty quantification. CP transforms a \emph{score function} into prediction sets that contain the true answer with high probability. While CP provides this coverage guarantee for arbitrary scores, the score quality significantly impacts prediction set sizes. Prior works have relied on LLM logits or other heuristic scores, lacking quality guarantees. We address this limitation by introducing CP-OPT, an optimization framework to learn scores that minimize set sizes while maintaining coverage. Furthermore, inspired by the Monty Hall problem, we extend CP's utility beyond uncertainty quantification to improve accuracy. We propose \emph{conformal revision of questions} (CROQ) to revise the problem by narrowing down the available choices to those in the prediction set. The coverage guarantee of CP ensures that the correct choice is in the revised question prompt with high probability, while the smaller number of choices increases the LLM's chances of answering it correctly. Experiments on MMLU, ToolAlpaca, and TruthfulQA datasets with Gemma-2, Llama-3 and Phi-3 models show that CP-OPT significantly reduces set sizes while maintaining coverage, and CROQ improves accuracy over the standard inference, especially when paired with CP-OPT scores. Together, CP-OPT and CROQ offer a robust framework for improving both the safety and accuracy of LLM-driven decision-making.

Personalized feedback plays a vital role in a student's learning process. While existing systems are adept at providing feedback over MCQ-based evaluation, this work focuses more on subjective and open-ended questions, which is similar to the problem of Automatic Short Answer Grading (ASAG) with feedback. Additionally, we introduce the Multimodal Short Answer grading with Feedback (MMSAF) problem over the traditional ASAG feedback problem to address the scenario where the student answer and reference answer might contain images. Moreover, we introduce the MMSAF dataset with 2197 data points along with an automated framework for generating such data sets. Our evaluations on existing LLMs over this dataset achieved an overall accuracy of 55\% on Level of Correctness labels, 75\% on Image Relevance labels and a score of 4.27 out of 5 in correctness level of LLM generated feedback as rated by experts. As per experts, Pixtral achieved a rating of above 4 out of all metrics, indicating that it is more aligned to human judgement, and that it is the best solution for assisting students.

Tuning the spatial extent of directional thermal emission across an arbitrary, and fixed spectral bandwidth is a fundamentally enabling capability for a range of emerging applications such as thermophotovoltaics, thermal imaging, and radiative cooling. However, previous experimental demonstrations were limited to narrow bandwidths, and the resonance frequency itself changed significantly as a function of the reconfigured directional response. Here, we demonstrate thermo-optic tuning of directional infrared emissivity using InAs-based gradient ENZ materials functioning as broadband directional thermal emitters whose angular selectivity can be modified via thermal free-carrier effects. We experimentally demonstrate two emitters achieving a 5{\deg} and 10{\deg} increase in the angular extent of their directional emissivity in the p-polarization across a prescribed, broad wavelength range of operation (12.5 to 15$\mu$m), for moderate temperatures below 400 K. Temperature-driven control of directional emissivity offers a new mode of post-fabrication control of radiative heat transfer that may in turn enable novel device functionalities., Comment: 16 pages, 5 figures

The success of deep learning in supervised fine-grained recognition for domain-specific tasks relies heavily on expert annotations. The Open-Set for fine-grained Self-Supervised Learning (SSL) problem aims to enhance performance on downstream tasks by strategically sampling a subset of images (the Core-Set) from a large pool of unlabeled data (the Open-Set). In this paper, we propose a novel method, BloomCoreset, that significantly reduces sampling time from Open-Set while preserving the quality of samples in the coreset. To achieve this, we utilize Bloom filters as an innovative hashing mechanism to store both low- and high-level features of the fine-grained dataset, as captured by Open-CLIP, in a space-efficient manner that enables rapid retrieval of the coreset from the Open-Set. To show the effectiveness of the sampled coreset, we integrate the proposed method into the state-of-the-art fine-grained SSL framework, SimCore [1]. The proposed algorithm drastically outperforms the sampling strategy of the baseline in SimCore [1] with a $98.5\%$ reduction in sampling time with a mere $0.83\%$ average trade-off in accuracy calculated across $11$ downstream datasets., Comment: Accepted at ICASSP 2025

Data scarcity in medical imaging poses significant challenges due to privacy concerns. Diffusion models, a recent generative modeling technique, offer a potential solution by generating synthetic and realistic data. However, questions remain about the performance of convolutional neural network (CNN) models on original and synthetic datasets. If diffusion-generated samples can help CNN models perform comparably to those trained on original datasets, reliance on patient-specific data for training CNNs might be reduced. In this study, we investigated the effectiveness of diffusion models for generating synthetic medical images to train CNNs in three domains: Brain Tumor MRI, Acute Lymphoblastic Leukemia (ALL), and SARS-CoV-2 CT scans. A diffusion model was trained to generate synthetic datasets for each domain. Pre-trained CNN architectures were then trained on these synthetic datasets and evaluated on unseen real data. All three datasets achieved promising classification performance using CNNs trained on synthetic data. Local Interpretable Model-Agnostic Explanations (LIME) analysis revealed that the models focused on relevant image features for classification. This study demonstrates the potential of diffusion models to generate synthetic medical images for training CNNs in medical image analysis.

How effective is peer-reviewing in identifying important papers? We treat this question as a forecasting task. Can we predict which papers will be highly cited in the future based on venue and "early returns" (citations soon after publication)? We show early returns are more predictive than venue. Finally, we end with constructive suggestions to address scaling challenges: (a) too many submissions and (b) too few qualified reviewers., Comment: The 31st International Conference on Computational Linguistics (COLING 2025)

A comprehensive study on persistent and thermonuclear burst emission of 4U 1728-34, commonly known as 'Slow Burster' is performed using seven archival observations of AstroSat spanning from 2016-2019. The burst-free persistent spectra can be well fitted with a blackbody (bbody) and a powerlaw (powerlaw) components, with a powerlaw photon index (Gamma) was found to be ~2 indicating the source was in "high/soft" bananna state or intermediate state. The time averaged power density spectrum reveals the presence of twin kilohertz Quasi Periodic Oscillations (kHz QPOs) with centroid frequencies 619+/-10 Hz and 965+/-6 Hz with a maximum fractional root mean squared amplitude of 6.24+/-1.31% at ~16 keV. From the upper kHz QPO, we infer the magnetospheric disk radius to be ~17 km, corresponding to a magnetic field strength of 0.35-1.27 * 10^7 G. The burst spectral evolution indicates Photospheric Radius Expansion (PRE) in five bursts, yeilding a touchdown radius of 3.1-5.47 km. These bursts reached near-Eddington luminosities, through which the distance of the source was calculated to be 5.18-5.21 kpc. Two of the bursts show coherent oscillations at 362.81-363.93 Hz. The presence of twin kHz QPOs and coherent Burst Oscillations allows us to provide two different estimates for the spin frequency of the Neutron Star in the system, for the first time using AstroSat., Comment: 18 pages, 13 Figures

We argue that the finiteness of quantum gravity amplitudes in fully compactified theories (at least in supersymmetric cases) leads to a bottom-up prediction for the existence of non-trivial dualities. In particular, finiteness requires the moduli space of massless fields to be compactifiable, meaning that its volume must be finite or at least grow no faster than that of Euclidean space. Moreover, we relate the compactifiability of moduli spaces to the condition that the lattice of charged objects transform in a semisimple representation under the action of the duality group. These ideas are supported by a wide variety of string theory examples., Comment: v2: 54 pages, 1 figure, references added, clarifications added

This work presents a comprehensive framework for the efficient implementation of finite-volume-based reacting flow solvers, specifically tailored for high speed propulsion applications. Using the exascale computing project (ECP) based AMReX framework, a compressible flow solver for handling high-speed reacting flows is developed. This work is complementary to the existing PeleC solver, emphasizing specific applications that include confined shock-containing flows, stationary and moving shocks and detonations. The framework begins with a detailed exposition of the numerical methods employed, emphasizing their application to complex geometries and their effectiveness in ensuring accurate and stable numerical simulations. Subsequently, an in-depth analysis evaluates the solver's performance across canonical and practical geometries, with particular focus on computational cost and efficiency. The solver's scalability and robustness are demonstrated through practical test cases, including flow path simulations of scramjet engines and detailed analysis of various detonation phenomena., Comment: 2024 12 01 V1

Neural radiance fields (NeRF) have exhibited highly photorealistic rendering of novel views through per-scene optimization over a single 3D scene. With the growing popularity of NeRF and its variants, they have become ubiquitous and have been identified as efficient 3D resources. However, they are still far from being scalable since a separate model needs to be stored for each scene, and the training time increases linearly with every newly added scene. Surprisingly, the idea of encoding multiple 3D scenes into a single NeRF model is heavily under-explored. In this work, we propose a novel conditional-cum-continual framework, called $C^{3}$-NeRF, to accommodate multiple scenes into the parameters of a single neural radiance field. Unlike conventional approaches that leverage feature extractors and pre-trained priors for scene conditioning, we use simple pseudo-scene labels to model multiple scenes in NeRF. Interestingly, we observe the framework is also inherently continual (via generative replay) with minimal, if not no, forgetting of the previously learned scenes. Consequently, the proposed framework adapts to multiple new scenes without necessarily accessing the old data. Through extensive qualitative and quantitative evaluation using synthetic and real datasets, we demonstrate the inherent capacity of the NeRF model to accommodate multiple scenes with high-quality novel-view renderings without adding additional parameters. We provide implementation details and dynamic visualizations of our results in the supplementary file.

Efficiently learning and executing long-horizon mobile manipulation (MoMa) tasks is crucial for advancing robotics in household and workplace settings. However, current MoMa models are data-inefficient, underscoring the need for improved models that require realistic-sized benchmarks to evaluate their efficiency, which do not exist. To address this, we introduce the LAMBDA ({\lambda}) benchmark (Long-horizon Actions for Mobile-manipulation Benchmarking of Directed Activities), which evaluates the data efficiency of models on language-conditioned, long-horizon, multi-room, multi-floor, pick-and-place tasks using a dataset of manageable size, more feasible for collection. The benchmark includes 571 human-collected demonstrations that provide realism and diversity in simulated and real-world settings. Unlike planner-generated data, these trajectories offer natural variability and replay-verifiability, ensuring robust learning and evaluation. We benchmark several models, including learning-based models and a neuro-symbolic modular approach combining foundation models with task and motion planning. Learning-based models show suboptimal success rates, even when leveraging pretrained weights, underscoring significant data inefficiencies. However, the neuro-symbolic approach performs significantly better while being more data efficient. Findings highlight the need for more data-efficient learning-based MoMa approaches. {\lambda} addresses this gap by serving as a key benchmark for evaluating the data efficiency of those future models in handling household robotics tasks., Comment: 8 pages

Matrix grammars are one of the first approaches ever proposed in regulated rewriting, prescribing that rules have to be applied in a certain order. Originally, they have been introduced by \'Abrah\'am on linguistic grounds. In traditional regulated rewriting, the most interesting case shows up when all rules are context-free. Typical descriptional complexity measures incorporate the number of nonterminals or the matrix length, i.e., the number of rules per matrix. When viewing matrices as program fragments, it becomes natural to consider additional applicability conditions for such matrices. Here, we focus on attaching a permitting and a forbidden string to every matrix in a matrix grammar. The matrix is applicable to a sentential form~$w$ only if the permitting string is a subword in~$w$ and the forbidden string is not a subword in~$w$. We call such a grammar, where the application of a matrix is conditioned as described, a semi-conditional matrix grammar. We consider $(1)$ the maximal lengths of permitting and forbidden strings, $(2)$ the number of nonterminals, $(3)$ the number of conditional matrices, $(4)$ the maximal length of any matrix and $(5)$ the number of conditional matrices with nonempty permitting and forbidden strings, as the resources (descriptional complexity measures) of a semi-conditional matrix grammar. In this paper, we show that certain semi-conditional matrix grammar families defined by restricting resources can generate all of the recursively enumerable languages.

Large-scale 2D datasets have been instrumental in advancing machine learning; however, progress in 3D vision tasks has been relatively slow. This disparity is largely due to the limited availability of 3D benchmarking datasets. In particular, creating real-world point cloud datasets for indoor scene semantic segmentation presents considerable challenges, including data collection within confined spaces and the costly, often inaccurate process of per-point labeling to generate ground truths. While synthetic datasets address some of these challenges, they often fail to replicate real-world conditions, particularly the occlusions that occur in point clouds collected from real environments. Existing 3D benchmarking datasets typically evaluate deep learning models under the assumption that training and test data are independently and identically distributed (IID), which affects the models' usability for real-world point cloud segmentation. To address these challenges, we introduce the BelHouse3D dataset, a new synthetic point cloud dataset designed for 3D indoor scene semantic segmentation. This dataset is constructed using real-world references from 32 houses in Belgium, ensuring that the synthetic data closely aligns with real-world conditions. Additionally, we include a test set with data occlusion to simulate out-of-distribution (OOD) scenarios, reflecting the occlusions commonly encountered in real-world point clouds. We evaluate popular point-based semantic segmentation methods using our OOD setting and present a benchmark. We believe that BelHouse3D and its OOD setting will advance research in 3D point cloud semantic segmentation for indoor scenes, providing valuable insights for the development of more generalizable models., Comment: 20 pages, 6 figures, 3 tables, accepted at ECCV 2024 Workshops

Contemporary embodied agents, such as Voyager in Minecraft, have demonstrated promising capabilities in open-ended individual learning. However, when powered with open large language models (LLMs), these agents often struggle with rudimentary tasks, even when fine-tuned on domain-specific knowledge. Inspired by human cultural learning, we present \collabvoyager, a novel framework that enhances Voyager with lifelong collaborative learning through explicit perspective-taking. \collabvoyager introduces three key innovations: (1) theory of mind representations linking percepts, beliefs, desires, and actions; (2) natural language communication between agents; and (3) semantic memory of task and environment knowledge and episodic memory of collaboration episodes. These advancements enable agents to reason about their and others' mental states, empirically addressing two prevalent failure modes: false beliefs and faulty task executions. In mixed-expertise Minecraft experiments, \collabvoyager agents outperform Voyager counterparts, significantly improving task completion rate by $66.6\% (+39.4\%)$ for collecting one block of dirt and $70.8\% (+20.8\%)$ for collecting one wood block. They exhibit emergent behaviors like knowledge transfer from expert to novice agents and collaborative code correction. \collabvoyager agents also demonstrate the ability to adapt to out-of-distribution tasks by using their previous experiences and beliefs obtained through collaboration. In this open-ended social learning paradigm, \collabvoyager paves the way for the democratic development of embodied AI, where agents learn in deployment from both peer and environmental feedback.

Miniaturization of electronic components has led to overheating, increasing power consumption and causing early circuit failures. Conventional heat dissipation methods are becoming inadequate due to limited surface area and higher short-circuit risks. This study presents a fast, low-cost, and scalable technique using 2D hexagonal boron nitride (hBN) coatings to enhance heat dissipation in commercial electronics. Inexpensive hBN layers, applied by drop casting or spray coating, boost thermal conductivity at IC surfaces from below 0.3 W/m-K to 260 W/m-K, resulting in over double the heat flux and convective heat transfer. This significantly reduces operating temperatures and power consumption, as demonstrated by a 17.4% reduction in a coated audio amplifier circuit board. Density functional theory indicates enhanced interaction between 2D hBN and packaging materials as a key factor. This approach promises substantial energy and cost savings for large-scale electronics without altering existing manufacturing processes., Comment: 27 Pages, 5 Figures

The exclusivity principle (E-principle) states that the sum of probabilities of pairwise exclusive events cannot exceed 1. Unlike other principles proposed to characterize quantum correlations, its intrinsically non-bipartite formulation enables its application in more general contextuality scenarios. Although equivalent to the no-signalling condition for any bipartite Bell scenario, this equivalence breaks down for two independent copies of the same scenario. Such violations of the E-principle due to multiple copies, known as its activation effect, was studied in [Nat Commun 4, 2263 (2013)] for the nonlocal extremal boxes of $(2,m,2)$, $(2,2,d)$, and $(3,2,2)$ Bell scenarios. The authors mapped the problem of exhibiting activation effects to finding certain cliques inside joint exclusivity graphs. In this work, we refine the joint exclusivity structure to be an edge-colored exclusivity multigraph. This allows us to draw a novel connection to Ramsey theory, which guarantees the existence of certain monochromatic subgraphs in sufficiently large edge-colored cliques, providing a powerful tool for ruling out E-principle violations. We then exploit this connection, drawing on Ramsey-theoretic results to study violations of the E-principle by multiple copies of the contextual extremal boxes of $n$-cycle scenarios, called $n$-cycle PR boxes, for the usual ($n=4$) PR box we show that the known E-principle violation of $5/4$ is the maximal achievable using two copies, and that this same upper bound applies to two copies of the KCBS ($n = 5$) PR box. We then prove that $n \geq 6$-cycle PR boxes, unlike the extremal boxes of the aforementioned Bell scenarios, do not exhibit activation effects with two or three copies. Finally, for any number of independent copies $k$, we establish a lower bound on $n$ above which $n$-cycle PR boxes do not exhibit activation effects with $k$ copies., Comment: 17 pages, 14 figures

In the high energy limit, $s\gg -t$, amplitudes in planar gauge theories Reggeize, with power law behavior $\big( \frac{s}{-t} \big)^{\alpha(t)}$ governed by the Regge trajectory $\alpha(t)$. Beyond the planar limit this simplicity is violated by "Regge cuts", for which practical organizational principles are still being developed. We use a top-down effective field theory organization based on color projection in the $t$ channel and rapidity evolution equations for collinear impact factors, to sum large $s\gg -t$ logarithms for Regge cut contributions. The results are matrix equations which are closed within a given color channel. To illustrate the method we derive in QCD with $SU(N_c)$ for the first time a closed 6$\times$6 evolution equation for the "decupletons" in the $\text{10}\oplus\overline{\text{10}}$ Regge color channel, a 2$\times$2 evolution equation for the "triantapentons" in the $\text{35}\oplus\overline{\text{35}}$ color channel, and a scalar evolution equation for the "tetrahexaconton" in the 64 color channel. More broadly, our approach allows us to describe generic Reggeization phenomena in non-planar gauge theories, providing valuable data for the all loop structure of amplitudes beyond the planar limit., Comment: 27 pages, 3 figures

Art has long been a medium for individuals to engage with the world. Scribble art, a form of abstract visual expression, features spontaneous, gestural strokes made with pens or brushes. These dynamic and expressive compositions, created quickly and impulsively, reveal intricate patterns and hidden meanings upon closer inspection. While scribble art is often associated with spontaneous expression and experimentation, it can also be planned and intentional. Some artists use scribble techniques as a starting point for their creative process, exploring the possibilities of line, shape, and texture before refining their work into more polished compositions. From ancient cave paintings to modern abstract sketches and doodles, scribble art has evolved with civilizations, reflecting diverse artistic movements and cultural influences. This evolution highlights its universal appeal, transcending language and cultural barriers and connecting people through the shared experience of creating art., Comment: SIGGRAPH Asia 2024

We explore various semantic understandings of dual intuitionistic logic by exploring the relationship between co-Heyting algebras and topological spaces. First, we discuss the relevant ideas in the setting of Heyting algebras and intuitionistic logic, showing organically the progression from the primordial example of lattices of open sets of topological spaces to more general ways of thinking about Heyting algebras. Then, we adapt the ideas to the dual intuitionistic setting, and use them to prove a number of interesting properties, including a deep relationship both intuitionistic and dual intuitionistic logic share through Kripke semantics to the modal logic $\mathsf{S4}$., Comment: Removed theorem 1.0.4 and proceeding discussion; report is clearer without that

Metrology, the science of measurement, plays a key role in Advanced Manufacturing (AM) to ensure quality control, process optimization, and predictive maintenance. However, it has often been overlooked in AM domains due to the current focus on automation and the complexity of integrated precise measurement systems. Over the years, Digital Twin (DT) technology in AM has gained much attention due to its potential to address these challenges through physical data integration and real-time monitoring, though its use in metrology remains limited. Taking this into account, this study proposes a novel framework, the Metrology and Manufacturing-Integrated Digital Twin (MM-DT), which focuses on data from two metrology tools, collected from Coordinate Measuring Machines (CMM) and FARO Arm devices. Throughout this process, we measured 20 manufacturing parts, with each part assessed twice under different temperature conditions. Using Ensemble Machine Learning methods, our proposed approach predicts measurement deviations accurately, achieving an R2 score of 0.91 and reducing the Root Mean Square Error (RMSE) to 1.59 micrometers. Our MM-DT framework demonstrates its efficiency by improving metrology processes and offers valuable insights for researchers and practitioners who aim to increase manufacturing precision and quality.

Large Language Models (LLMs) are highly resource-intensive to fine-tune due to their enormous size. While low-rank adaptation is a prominent parameter-efficient fine-tuning approach, it suffers from sensitivity to hyperparameter choices, leading to instability in model performance on fine-tuning downstream tasks. This paper highlights the importance of effective parameterization in low-rank fine-tuning to reduce estimator variance and enhance the stability of final model outputs. We propose MonteCLoRA, an efficient fine-tuning technique, employing Monte Carlo estimation to learn an unbiased posterior estimation of low-rank parameters with low expected variance, which stabilizes fine-tuned LLMs with only O(1) additional parameters. MonteCLoRA shows significant improvements in accuracy and robustness, achieving up to 3.8% higher accuracy and 8.6% greater robustness than existing efficient fine-tuning methods on natural language understanding tasks with pre-trained RoBERTa-base. Furthermore, in generative tasks with pre-trained LLaMA-1-7B, MonteCLoRA demonstrates robust zero-shot performance with 50% lower variance than the contemporary efficient fine-tuning methods. The theoretical and empirical results presented in the paper underscore how parameterization and hyperpriors balance exploration-exploitation in the low-rank parametric space, therefore leading to more optimal and robust parameter estimation during efficient fine-tuning., Comment: 48 pages, 10 figures, 10 tables, Code: https://github.com/LCS2-IIITD/MonteCLoRA

Data scarcity and distribution shifts often hinder the ability of machine learning models to generalize when applied to proteins and other biological data. Self-supervised pre-training on large datasets is a common method to enhance generalization. However, striving to perform well on all possible proteins can limit model's capacity to excel on any specific one, even though practitioners are often most interested in accurate predictions for the individual protein they study. To address this limitation, we propose an orthogonal approach to achieve generalization. Building on the prevalence of self-supervised pre-training, we introduce a method for self-supervised fine-tuning at test time, allowing models to adapt to the test protein of interest on the fly and without requiring any additional data. We study our test-time training (TTT) method through the lens of perplexity minimization and show that it consistently enhances generalization across different models, their scales, and datasets. Notably, our method leads to new state-of-the-art results on the standard benchmark for protein fitness prediction, improves protein structure prediction for challenging targets, and enhances function prediction accuracy.

We consider the problem of multiclass transductive online learning when the number of labels can be unbounded. Previous works by Ben-David et al. [1997] and Hanneke et al. [2023b] only consider the case of binary and finite label spaces, respectively. The latter work determined that their techniques fail to extend to the case of unbounded label spaces, and they pose the question of characterizing the optimal mistake bound for unbounded label spaces. We answer this question by showing that a new dimension, termed the Level-constrained Littlestone dimension, characterizes online learnability in this setting. Along the way, we show that the trichotomy of possible minimax rates of the expected number of mistakes established by Hanneke et al. [2023b] for finite label spaces in the realizable setting continues to hold even when the label space is unbounded. In particular, if the learner plays for $T \in \mathbb{N}$ rounds, its minimax expected number of mistakes can only grow like $\Theta(T)$, $\Theta(\log T)$, or $\Theta(1)$. To prove this result, we give another combinatorial dimension, termed the Level-constrained Branching dimension, and show that its finiteness characterizes constant minimax expected mistake-bounds. The trichotomy is then determined by a combination of the Level-constrained Littlestone and Branching dimensions. Quantitatively, our upper bounds improve upon existing multiclass upper bounds in Hanneke et al. [2023b] by removing the dependence on the label set size. In doing so, we explicitly construct learning algorithms that can handle extremely large or unbounded label spaces. A key and novel component of our algorithm is a new notion of shattering that exploits the sequential nature of transductive online learning. Finally, we complete our results by proving expected regret bounds in the agnostic setting, extending the result of Hanneke et al. [2023b]., Comment: 23 pages

Over the years, Digital Twin (DT) has become popular in Advanced Manufacturing (AM) due to its ability to improve production efficiency and quality. By creating virtual replicas of physical assets, DTs help in real-time monitoring, develop predictive models, and improve operational performance. However, integrating data from physical systems into reliable predictive models, particularly in precision measurement and failure prevention, is often challenging and less explored. This study introduces a Predictive Maintenance and Inspection Digital Twin (PMI-DT) framework with a focus on precision measurement and predictive quality assurance using 3D-printed 1''-4 ACME bolt, CyberGage 360 vision inspection system, SolidWorks, and Microsoft Azure. During this approach, dimensional inspection data is combined with fatigue test results to create a model for detecting failures. Using Machine Learning (ML) -- Random Forest and Decision Tree models -- the proposed approaches were able to predict bolt failure with real-time data 100% accurately. Our preliminary result shows Max Position (30%) and Max Load (24%) are the main factors that contribute to that failure. We expect the PMI-DT framework will reduce inspection time and improve predictive maintenance, ultimately giving manufacturers a practical way to boost product quality and reliability using DT in AM.

We study learning in a dynamically evolving environment modeled as a Markov game between a learner and a strategic opponent that can adapt to the learner's strategies. While most existing works in Markov games focus on external regret as the learning objective, external regret becomes inadequate when the adversaries are adaptive. In this work, we focus on \emph{policy regret} -- a counterfactual notion that aims to compete with the return that would have been attained if the learner had followed the best fixed sequence of policy, in hindsight. We show that if the opponent has unbounded memory or if it is non-stationary, then sample-efficient learning is not possible. For memory-bounded and stationary, we show that learning is still statistically hard if the set of feasible strategies for the learner is exponentially large. To guarantee learnability, we introduce a new notion of \emph{consistent} adaptive adversaries, wherein, the adversary responds similarly to similar strategies of the learner. We provide algorithms that achieve $\sqrt{T}$ policy regret against memory-bounded, stationary, and consistent adversaries., Comment: NeurIPS'24; fix typos