We introduce an uncertainty-aware graph representation framework for learning to guide planning in Partially Observable Markov Decision Processes (POMDPs). Unlike existing approaches that require domain or problem size specific neural architectures, GammaZero leverages a unified graph-based belief representation that enables generalization across problem sizes within a domain. Our key insight is that belief states can be systematically transformed into uncertainty-aware graphs where structural patterns learned on small problems transfer to larger instances. We employ a graph neural network with a decoder architecture to learn value functions and policies from expert demonstrations on computationally tractable problems, then apply these learned heuristics to guide Monte Carlo tree search on larger problems. Experimental results on standard POMDP benchmarks demonstrate that GammaZero achieves comparable performance to BetaZero when trained and tested on the same-sized problems, while enabling zero-shot generalization to problems 2-6× larger than those seen during training.

Developers spend substantial time searching and navigating unfamiliar local codebases, often incurring significant cognitive overhead during tasks such as feature location and debugging. Traditional information retrieval methods rely on keyword matching and struggle with the vocabulary mismatch problem between natural language queries and source code. Although deep learning approaches support semantic code search, they typically require large training datasets and may not generalize well to local repositories. Pre-trained programming language models face similar limitations because they are not optimized for project-specific conventions and structure. To address these challenges, we present the Learning to Navigate framework, which models local code search as guided traversal over a repository graph using beam search with a learnable heuristic. We instantiate the framework with a composite graph that integrates hierarchical structure and function call relationships and evaluate it on six popular Python repositories. Our approach achieves an average Mean Reciprocal Rank (MRR) of 0.91 and Recall@1 (R@1) of 0.86, outperforming a fine-tuned UniXCoder baseline. In addition to improved retrieval accuracy, the framework surfaces structured navigation paths that support adaptive, interpretable code exploration.

Surgical procedures such as laparoscopic appendectomy are complex, high-stakes processes, yet formalizing their workflows for decision support remains a significant challenge. Inducing probabilistic planning domain models in this setting is particularly difficult due to the lack of structured event data and the prevalence of implicit actions in clinical narratives, which neither empirical symbolic methods nor Large Language Models (LLMs) can adequately address on their own. We introduce NSPIN, a neurosymbolic framework for inducing probabilistic planning domain models from unstructured clinical narratives. Our method extracts and imputes structured event sequences from raw text using a pretrained LLM, then induces a PPDDL model and refines its preconditions with LLM-proposed revisions, guided by empirical validation. We evaluate the approach on 2,660 laparoscopic appendectomy notes written by 9 surgeons. NSPIN yields models that generalize to unseen notes, and expert clinical review indicates its induced knowledge is largely consistent with surgical practice.

Translating natural-language planning intent into verified plans is a longstanding challenge: people communicate goals in language, while classical planners require formal PDDL specifications. Recent agentic frameworks bridge this gap by orchestrating a pool of specialized repair agents inside a verifier-checked refinement loop, but the orchestrator at the centre is itself a prompted frontier LLM, paying a frontier-LLM API call at every refinement step. We present HALO (Hybrid Agent-Learned Orchestrator), which trains the orchestrator from refinement trajectories that an external verifier has certified as ending in valid plans, across 11 PDDL domains. HALO pairs a small QLoRA-tuned policy with three hardcoded rules for trivially decidable selections, and operates over an expanded 21-agent action space. Unlike approaches that prompt a frontier LLM at every step or learn an orchestrator from sparse end-of-episode rewards, our key observation is that the verifier already provides strong guidance: every accepted trajectory is a sequence of demonstrably correct (state, agent) decisions, directly usable as supervision. Across PlanBench, Natural Plan, and classical planning benchmarks, HALO matches or exceeds the GPT-5-mini prompted baseline on success rate, sits within three percentage points of the stronger Gemini-3-Flash prompted baseline, reduces orchestration cost by more than an order of magnitude ($0.18 to $0.004 per task against GPT-5-mini, roughly 45× cheaper; roughly 15× cheaper than Gemini-3-Flash), and cuts total LLM calls per episode by 40 to 50 percent.

While deep neural networks are extremely effective at classifying images, they remain opaque and hard to interpret. We introduce local and global explanation methods for black-box models that generate explanations in terms of human-recognizable primitive concepts. Both the local explanations for a single image and the global explanations for a set of images are cast as logical formulas in monotone disjunctive-normal-form (MDNF), whose satisfaction guarantees that the model yields a high score on a given class. We also present an algorithm for explaining the classification of examples into multiple classes in the form of a monotone explanation list over primitive concepts. Despite their simplicity and interpretability we show that the explanations maintain high fidelity and coverage with respect to the blackbox models they seek to explain in challenging vision datasets.

Multiagent Reinforcement Learning (MARL) poses significant challenges due to the exponential growth of state and action spaces and the non-stationary nature of multiagent environments. This results in notable sample inefficiency and hinders generalization across diverse tasks. The complexity is further pronounced in relational settings, where domain knowledge is crucial but often underutilized by existing MARL algorithms. To overcome these hurdles, we propose integrating relational planners as centralized controllers with efficient state abstractions and reinforcement learning. This approach proves to be sample-efficient and facilitates effective task transfer and generalization.

Current chess rating systems update ratings incrementally and may not always accurately reflect a player's true strength at all times, especially for rapidly improving players or very rusty players. To overcome this, we explore a method to estimate player ratings directly from game moves and clock times. We compiled a benchmark dataset from Lichess with over one million games, encompassing various time controls and including move sequences and clock times. Our model architecture comprises a CNN to learn positional features, which are then integrated with clock-time data into a Bidirectional LSTM, predicting player ratings after each move. The model achieved an MAE of 182 rating points on the test data. Additionally, we applied our model to the 2024 IEEE Big Data Cup Chess Puzzle Difficulty Competition dataset, predicted puzzle ratings and achieved competitive results. This model is the first to use no hand-crafted features to estimate chess ratings and also the first to output a rating prediction after each move. Our method highlights the potential of using move-based rating estimation for enhancing rating systems and potentially other applications such as cheating detection.

Explaining the decisions of learned models in human-interpretable language is critical for building trustworthy AI. Much of the current work on explainable AI in image classification focuses on generating local (image-specific) explanations that highlight parts of the image responsible for decisions. Although valuable, these local explanations are often not human-interpretable, lack generalizability across entire classes or datasets, and fail to cater to a diverse set of stakeholders. In this work, we introduce a novel framework for generating global explanations in terms of human-aligned concepts that are applicable to any image-based classifier, irrespective of the architecture. Our methods provide both local and global explanations across multiple images from multiple classes. We present our framework for generating global explanations along with experimental results on multiple datasets that demonstrate the effectiveness of our technique. Our method achieves a test coverage of 99.3% on the Stanford Cars dataset.

We introduce an action-centric graph representation framework for learning to guide planning in Partially Observable Markov Decision Processes (POMDPs). Unlike existing approaches that require domain-specific neural architectures and struggle with scalability, GammaZero leverages a unified graph-based belief representation that enables generalization across problem sizes within a domain. Our key insight is that belief states can be systematically transformed into action-centric graphs where structural patterns learned on small problems transfer to larger instances. We employ a graph neural network with a decoder architecture to learn value functions and policies from expert demonstrations on computationally tractable problems, then apply these learned heuristics to guide Monte Carlo tree search on larger problems. Experimental results on standard POMDP benchmarks demonstrate that GammaZero achieves comparable performance to BetaZero when trained and tested on the same-sized problems, while uniquely enabling zero-shot generalization to problems 2-4 times larger than those seen during training, maintaining solution quality with reduced search requirements.

Missing values in multivariate time series data can harm machine learning performance and introduce bias. These gaps arise from sensor malfunctions, blackouts, and human error and are typically addressed by data imputation. Previous work has tackled the imputation of missing data in random, complete blackouts and forecasting scenarios. The current paper addresses a more general missing pattern, which we call "partial blackout," where a subset of features is missing for consecutive time steps. We introduce a two-stage imputation process using self-attention and diffusion processes to model feature and temporal correlations. Notably, our model effectively handles missing data during training, enhancing adaptability and ensuring reliable imputation and performance, even with incomplete datasets. Our experiments on benchmark and two real-world time series datasets demonstrate that our model outperforms the state-of-the-art in partial blackout scenarios and shows better scalability.

Deep learning models are notoriously opaque. Existing explanation methods often focus on localized visual explanations for individual images. Concept-based explanations, while offering global insights, require extensive annotations, incurring significant labeling cost. We propose an approach that leverages user-defined part labels from a limited set of images and efficiently transfers them to a larger dataset. This enables the generation of global symbolic explanations by aggregating part-based local explanations, ultimately providing human-understandable explanations for model decisions on a large scale.

We propose a novel approach to learn relational policies for classical planning based on learning to rank actions. We introduce a new graph representation that explicitly captures action information and propose a Graph Neural Network (GNN) architecture augmented with Gated Recurrent Units (GRUs) to learn action rankings. Unlike value-function based approaches that must learn a globally consistent function, our action ranking method only needs to learn locally consistent ranking. Our model is trained on data generated from small problem instances that are easily solved by planners and is applied to significantly larger instances where planning is computationally prohibitive. Experimental results across standard planning benchmarks demonstrate that our action-ranking approach not only achieves better generalization to larger problems than those used in training but also outperforms multiple baselines (value function and action ranking) methods in terms of success rate and plan quality.

Beam search decoding is the de-facto method for decoding auto-regressive Neural Machine Translation (NMT) models, including multilingual NMT where the target language is specified as an input. However, decoding multilingual NMT models commonly produces "offtarget" translations - yielding translation outputs not in the intended language. In this paper, we first conduct an error analysis of off-target translations for a strong multilingual NMT model and identify how these decodings are produced during beam search. We then propose Language-informed Beam Search (LiBS), a general decoding algorithm incorporating an off-the-shelf Language Identification (LiD) model into beam search decoding to reduce off-target translations. LiBS is an inferencetime procedure that is NMT-model agnostic and does not require any additional parallel data. Results show that our proposed LiBS algorithm on average improves +1.1 BLEU and +0.9 BLEU on WMT and OPUS datasets, and reduces off-target rates from 22.9% to 7.7% and 65.8% to 25.3% respectively.

We present an online planning framework and a new benchmark dataset for solving multi-object rearrangement problems in partially observable, multi-room environments. Current object rearrangement solutions, primarily based on Reinforcement Learning or hand-coded planning methods, often lack adaptability to diverse challenges. To address this limitation, we introduce a novel Hierarchical Object-Oriented Partially Observed Markov Decision Process (HOO-POMDP) planning approach. This approach comprises of (a) an object-oriented POMDP planner generating sub-goals, (b) a set of low-level policies for sub-goal achievement, and (c) an abstraction system converting the continuous low-level world into a representation suitable for abstract planning. To enable rigorous evaluation of rearrangement challenges, we introduce MultiRoomR, a comprehensive benchmark featuring diverse multi-room environments with varying degrees of partial observability (10-30% initial visibility), blocked paths, obstructed goals, and multiple objects (10-20) distributed across 2-4 rooms. Experiments demonstrate that our system effectively handles these complex scenarios while maintaining robust performance even with imperfect perception, achieving promising results across both existing benchmarks and our new MultiRoomR dataset.

Time-series modeling is essential for applications in agriculture, weather forecasting, food production, and more. However, missing data due to sensor malfunctions, power outages, and human errors is a common issue, complicating the training of machine learning models. We propose a diffusion-based generative model to address this problem and fill the gaps in the data. Our approach captures feature and time correlations through a two-stage imputation process. Our model outperforms state-of-the-art imputation methods and is more scalable in GPU resources.

We study reward poisoning attacks on Combinatorial Multi-armed Bandits (CMAB). We first provide a sufficient and necessary condition for the attackability of CMAB, a notion to capture the vulnerability and robustness of CMAB. The attackability condition depends on the intrinsic properties of the corresponding CMAB instance such as the reward distributions of super arms and outcome distributions of base arms. Additionally, we devise an attack algorithm for attackable CMAB instances. Contrary to prior understanding of multi-armed bandits, our work reveals a surprising fact that the attackability of a specific CMAB instance also depends on whether the bandit instance is known or unknown to the adversary. This finding indicates that adversarial attacks on CMAB are difficult in practice and a general attack strategy for any CMAB instance does not exist since the environment is mostly unknown to the adversary. We validate our theoretical findings via extensive experiments on real-world CMAB applications including probabilistic maximum covering problem, online minimum spanning tree, cascading bandits for online ranking, and online shortest path.

We propose a hybrid planner-(deep)reinforcement learning (RL) architecture, RePReL, that leverages a relational planner to efficiently provide useful state abstractions. State abstractions have a tremendous advantage for better generalization and transfer in RL. Our framework takes an important step toward constructing these abstractions. Specifically, the framework enables multi-level abstractions by leveraging a high-level planner to communicate with a low-level (deep) reinforcement learner. Our empirical results demonstrate the generalization and transfer capabilities of the framework in both discrete and continuous domains with rich structures (objects and relations between these objects). A key aspect of RePReL is that it can be seen as a plug-and-play framework where different planners can be used in combination with different (deep) RL agents.

We hypothesize that deep network classifications of complex scenes can be explained using sets of relevant objects. We employ beam search and singular value decomposition to generate local and global explanations that summarize the deep model's interpretation of a class.

Ensemble models (bagging and gradient-boosting) of relational decision trees have proved to be some of the most effective learning methods in the area of probabilistic logic models (PLMs). While effective, they lose one of the most important benefits of PLMs—interpretability. In this paper we consider the problem of compressing a large set of learned trees into a single explainable model. To this effect, we propose CoTE—Compression of Tree Ensembles—that produces a single small decision list as a compressed representation. CoTE first converts the trees to decision lists and then performs the combination and compression with the aid of the original training set. An experimental evaluation demonstrates the effectiveness of CoTE in several benchmark relational data sets.