Perception and reasoning are two representative abilities of intelligence that are integrated seamlessly during human problem-solving processes.

Since the advent of large language models (LLMs), research has focused on instruction following and deductive reasoning. A central question remains: can these models discover new knowledge, and how can we evaluate this ability? We address this by studying abductive reasoning-the generation of plausible hypotheses to explain observations-and introduce GEAR (General Evaluation for Abductive Reasoning), a general-purpose, fully automated, transparent, and label-free evaluation paradigm. GEAR scores hypothesis sets by three metrics: consistency (each hypothesis explains the observations), generalizability (consistent hypotheses make meaningful predictions on unseen inputs), and diversity (the set covers distinct predictions and patterns). Built this way, GEAR is scalable (no human gold answers), reliable (deterministic scoring aligned with classical abduction), and open-ended (scores improve only when models produce new plausible hypotheses, unlike static benchmarks that saturate once accuracy is high). Using GEAR, we conduct a fine-grained study of nine LLMs on four abduction benchmarks with 1,500 problems, generating over 50,000 candidate hypotheses and revealing model differences obscured by gold-answer or purely human evaluations. We further propose a momentum-based curriculum that adjusts GEAR-derived training data by learning velocity: it starts with what the model learns quickly and shifts toward harder objectives such as generating diverse hypotheses once the model is confident on foundational objectives. Without gold-label supervision, this strategy improves all GEAR objectives and these gains transfer to established abductive reasoning benchmarks. Taken together, GEAR provides a principled framework that evaluates abduction and supplies label-free, scalable training signals that help LLMs produce more diverse and reliable hypotheses.

The capability to store data about business processes execution in so-called Event Logs has brought to the diffusion of tools for the analysis of process executions and for the assessment of the goodness of a process model. Nonetheless, these tools are often very rigid in dealing with with Event Logs that include incomplete information about the process execution. Thus, while the ability of handling incomplete event data is one of the challenges mentioned in the process mining manifesto, the evaluation of compliance of an execution trace still requires an end-to-end complete trace to be performed. This paper exploits the power of abduction to provide a flexible, yet computationally effective, framework to deal with different forms of incompleteness in an Event Log. Moreover it proposes a refinement of the classical notion of compliance into strong and conditional compliance to take into account incomplete logs. Finally, performances evaluation in an experimental setting shows the feasibility of the presented approach.

Abductive reasoning, reasoning for inferring explanations for observations, is often mentioned in scientific, design-related and artistic contexts, but its understanding varies across these domains. This paper reviews how abductive reasoning is discussed in epistemology, science and design, and then analyses how various computational systems use abductive reasoning. Our analysis shows that neither theoretical accounts nor computational implementations of abductive reasoning adequately address generating creative hypotheses. Theoretical frameworks do not provide a straightforward model for generating creative abductive hypotheses, and computational systems largely implement syllogistic forms of abductive reasoning. We break down abduc-tive computational systems into components and conclude by identifying specific directions for future research that could advance the state of creative abductive reasoning in computational systems.

Pearl observes that causal knowledge enables predicting the effects of interventions, such as actions, whereas descriptive knowledge only permits drawing conclusions from observation. This paper extends Pearl's approach to causality and interventions to the setting of stratified abductive logic programs. It shows how stable models of such programs can be given a causal interpretation by building on philosophical foundations and recent work by Bochman and Eelink et al. In particular, it provides a translation of abductive logic programs into causal systems, thereby clarifying the informal causal reading of logic program rules and supporting principled reasoning about external actions. The main result establishes that the stable model semantics for stratified programs conforms to key philosophical principles of causation, such as causal sufficiency, natural necessity, and irrelevance of unobserved effects. This justifies the use of stratified abductive logic programs as a framework for causal modeling and for predicting the effects of interventions.

Abductive reasoning in knowledge graphs aims to generate plausible logical hypotheses from observed entities, with broad applications in areas such as clinical diagnosis and scientific discovery. However, due to a lack of controllability, a single observation may yield numerous plausible but redundant or irrelevant hypotheses on large-scale knowledge graphs. To address this limitation, we introduce the task of controllable hypothesis generation to improve the practical utility of abductive reasoning. This task faces two key challenges when controlling for generating long and complex logical hypotheses: hypothesis space collapse and hypothesis oversensitivity. To address these challenges, we propose CtrlHGen, a Controllable logcial Hypothesis Generation framework for abductive reasoning over knowledge graphs, trained in a two-stage paradigm including supervised learning and subsequent reinforcement learning. To mitigate hypothesis space collapse, we design a dataset augmentation strategy based on sub-logical decomposition, enabling the model to learn complex logical structures by leveraging semantic patterns in simpler components. To address hypothesis oversensitivity, we incorporate smoothed semantic rewards including Dice and Overlap scores, and introduce a condition-adherence reward to guide the generation toward user-specified control constraints. Extensive experiments on three benchmark datasets demonstrate that our model not only better adheres to control conditions but also achieves superior semantic similarity performance compared to baselines.

Significant research efforts have been made to scale and improve vision-language model (VLM) training approaches. Yet, with an ever-growing number of benchmarks,researchers are tasked with the heavy burden of implementing each protocol, bearing a non-trivial computational cost, and making sense of how all these benchmarks translate into meaningful axes of progress.To facilitate a systematic evaluation of VLM progress, we introduce UniBench: a unified implementation of 50 VLM benchmarks spanning a range of carefully categorized vision-centric capabilities from object recognition to spatial awareness, counting, and much more. We showcase the utility of UniBench for measuring progress by evaluating nearly 60 publicly available vision-language models, trained on scales of up to 12.8B samples. We find that while scaling training data or model size can boost many vision-language model capabilities, scaling offers little benefit for reasoning or relations. Surprisingly, we also discover today's best VLMs struggle on simple digit recognition and counting tasks, e.g.

Logical Credal Networks or LCNs were recently introduced as a powerful probabilistic logic framework for representing and reasoning with imprecise knowledge. Unlike many existing formalisms, LCNs have the ability to represent cycles and allow specifying marginal and conditional probability bounds on logic formulae which may be important in many realistic scenarios. Previous work on LCNs has focused exclusively on marginal inference, namely computing posterior lower and upper probability bounds on a query formula. In this paper, we explore abductive reasoning tasks such as solving MAP and Marginal MAP queries in LCNs given some evidence. We first formally define the MAP and Marginal MAP tasks for LCNs and subsequently show how to solve these tasks exactly using search-based approaches.