Agents capable of accomplishing complex tasks through multiple interactions with the environment have emerged as a popular research direction. However, in such multi-step settings, the conventional group-level policy optimization algorithm becomes suboptimal because of its underlying assumption that each action holds equal contribution, which deviates significantly from reality. Our analysis reveals that only a small fraction of actions are critical in determining the final outcome. Building on this insight, we propose CARL, a critical-action-focused reinforcement learning algorithm tailored for multi-step agents. CARL achieves focused training through providing action-level optimization signals for high-criticality actions while excluding low-criticality actions from model update. Extensive experiments demonstrate that CARL achieves both stronger performance and higher efficiency during training and inference across diverse evaluation settings.

The paper introduces a novel behavioural translation style space (BTSS) that describes possible behavioural translation patterns. The suggested BTSS is organized as a hierarchical structure that entails various embedded processing layers. We posit that observable translation behaviour - i.e. eye and finger movements - is fundamental when executing the physical act of translation but it is caused and shaped by higher-order cognitive processes and affective translation states. We analyse records of keystrokes and gaze data as indicators of the hidden mental processing structure and organize the behavioural patterns as a multi-layered embedded BTSS. We develop a perspective in which the BTSS serves as the basis for a computational translation agent to simulate the temporal dynamics of affect, behavioural routines and cognition during human translation production.

Spectral imaging offers promising applications across diverse domains, including medicine and urban scene understanding, and is already established as a critical modality in remote sensing. However, variability in channel dimensionality and captured wavelengths among spectral cameras impede the development of AI-driven methodologies, leading to camera-specific models with limited generalizability and inadequate cross-camera applicability. To address this bottleneck, we introduce CARL, a model for Camera-Agnostic Representation Learning across RGB, multispectral, and hyperspectral imaging modalities. To enable the conversion of a spectral image with any channel dimensionality to a camera-agnostic representation, we introduce a novel spectral encoder, featuring a self-attention-cross-attention mechanism, to distill salient spectral information into learned spectral representations. Spatio-spectral pre-training is achieved with a novel feature-based self-supervision strategy tailored to CARL. Large-scale experiments across the domains of medical imaging, autonomous driving, and satellite imaging demonstrate our model's unique robustness to spectral heterogeneity, outperforming on datasets with simulated and real-world cross-camera spectral variations. The scalability and versatility of the proposed approach position our model as a backbone for future spectral foundation models.

Constrained Stochastic Shortest Path Problems (CSSPs) model problems with probabilistic effects, where a primary cost is min-imised subject to constraints over secondary costs, e.g., minimise time subject to monetary budget. Current heuristic search algorithms for CSSPs solve a sequence of increasingly larger CSSPs as linear programs until an optimal solution for the original CSSP is found. In this paper, we introduce a novel algorithm CARL, which solves a series of unconstrained Stochastic Shortest Path Problems (SSPs) with efficient heuristic search algorithms. These SSP subproblems are constructed with scalarisations that project the CSSP's vector of primary and secondary costs onto a scalar cost. CARL finds a maximising scalarisation using an optimisation algorithm similar to the subgradient method which, together with the solution to its associated SSP, yields a set of policies that are combined into an optimal policy for the CSSP . Our experiments show that CARL solves 50% more problems than the state-of-the-art on existing benchmarks.

Video synchronization-aligning multiple video streams capturing the same event from different angles-is crucial for applications such as reality TV show production, sports analysis, surveillance, and autonomous systems. Prior work has heavily relied on audio cues or specific visual events, limiting applicability in diverse settings where such signals may be unreliable or absent. Additionally, existing benchmarks for video synchronization lack generality and reproducibility, restricting progress in the field. In this work, we introduce VideoSync, a video synchronization framework that operates independently of specific feature extraction methods, such as human pose estimation, enabling broader applicability across different content types. We evaluate our system on newly composed datasets covering single-human, multi-human, and non-human scenarios, providing both the methodology and code for dataset creation to establish reproducible benchmarks. Our analysis reveals biases in prior SOTA work, particularly in SeSyn-Net's preprocessing pipeline, leading to inflated performance claims. We correct these biases and propose a more rigorous evaluation framework, demonstrating that VideoSync outperforms existing approaches, including SeSyn-Net, under fair experimental conditions. Additionally, we explore various synchronization offset prediction methods, identifying a convolutional neural network (CNN)-based model as the most effective. Our findings advance video synchronization beyond domain-specific constraints, making it more generalizable and robust for real-world applications.

Performance predictors have emerged as a promising method to accelerate the evaluation stage of neural architecture search (NAS). These predictors estimate the performance of unseen architectures by learning from the correlation between a small set of trained architectures and their performance. However, most existing predictors ignore the inherent distribution shift between limited training samples and diverse test samples. Hence, they tend to learn spurious correlations as shortcuts to predictions, leading to poor generalization. To address this, we propose a Causality-guided Architecture Representation Learning (CARL) method aiming to separate critical (causal) and redundant (non-causal) features of architectures for generalizable architecture performance prediction. Specifically, we employ a substructure extractor to split the input architecture into critical and redundant substructures in the latent space. Then, we generate multiple interventional samples by pairing critical representations with diverse redundant representations to prioritize critical features. Extensive experiments on five NAS search spaces demonstrate the state-of-the-art accuracy and superior interpretability of CARL. For instance, CARL achieves 97.67% top-1 accuracy on CIFAR-10 using DARTS.

--While pre-trained language models excel at semantic understanding, they often struggle to capture nuanced affective information critical for affective recognition tasks. T o address these limitations, we propose a novel framework for enhancing emotion-aware embeddings in transformer-based models. Our approach introduces a continuous valence-arousal labeling system to guide contrastive learning, which captures subtle and multidimensional emotional nuances more effectively. Furthermore, we employ a dynamic token perturbation mechanism, using gradient-based saliency to focus on sentiment-relevant tokens, improving model sensitivity to emotional cues. The experimental results demonstrate that the proposed framework outperforms existing methods, achieving up to 15.5% improvement in the emotion classification benchmark, highlighting the importance of employing continuous labels. This improvement demonstrates that the proposed framework is effective in affective representation learning and enables precise and contextually relevant emotional understanding.

Cooperative Adaptive Cruise Control (CACC) plays a pivotal role in enhancing traffic efficiency and safety in Connected and Autonomous Vehicles (CAVs). Reinforcement Learning (RL) has proven effective in optimizing complex decision-making processes in CACC, leading to improved system performance and adaptability. Among RL approaches, Multi-Agent Reinforcement Learning (MARL) has shown remarkable potential by enabling coordinated actions among multiple CAVs through Centralized Training with Decentralized Execution (CTDE). However, MARL often faces scalability issues, particularly when CACC vehicles suddenly join or leave the platoon, resulting in performance degradation. To address these challenges, we propose Communication-Aware Reinforcement Learning (CA-RL). CA-RL includes a communication-aware module that extracts and compresses vehicle communication information through forward and backward information transmission modules. This enables efficient cyclic information propagation within the CACC traffic flow, ensuring policy consistency and mitigating the scalability problems of MARL in CACC. Experimental results demonstrate that CA-RL significantly outperforms baseline methods in various traffic scenarios, achieving superior scalability, robustness, and overall system performance while maintaining reliable performance despite changes in the number of participating vehicles.

Modern large-scale recommender systems are built upon computation-intensive infrastructure and usually suffer from a huge difference in traffic between peak and off-peak periods. In peak periods, it is challenging to perform real-time computation for each request due to the limited budget of computational resources. The recommendation with a cache is a solution to this problem, where a user-wise result cache is used to provide recommendations when the recommender system cannot afford a real-time computation. However, the cached recommendations are usually suboptimal compared to real-time computation, and it is challenging to determine the items in the cache for each user. In this paper, we provide a cache-aware reinforcement learning (CARL) method to jointly optimize the recommendation by real-time computation and by the cache. We formulate the problem as a Markov decision process with user states and a cache state, where the cache state represents whether the recommender system performs recommendations by real-time computation or by the cache. The computational load of the recommender system determines the cache state. We perform reinforcement learning based on such a model to improve user engagement over multiple requests. Moreover, we show that the cache will introduce a challenge called critic dependency, which deteriorates the performance of reinforcement learning. To tackle this challenge, we propose an eigenfunction learning (EL) method to learn independent critics for CARL. Experiments show that CARL can significantly improve the users' engagement when considering the result cache. CARL has been fully launched in Kwai app, serving over 100 million users.

The generation of comprehensible explanations is an essential feature of modern artificial intelligence systems. In this work, we consider probabilistic logic programming, an extension of logic programming which can be useful to model domains with relational structure and uncertainty. Essentially, a program specifies a probability distribution over possible worlds (i.e., sets of facts). The notion of explanation is typically associated with that of a world, so that one often looks for the most probable world as well as for the worlds where the query is true. Unfortunately, such explanations exhibit no causal structure. In particular, the chain of inferences required for a specific prediction (represented by a query) is not shown. In this paper, we propose a novel approach where explanations are represented as programs that are generated from a given query by a number of unfolding-like transformations. Here, the chain of inferences that proves a given query is made explicit. Furthermore, the generated explanations are minimal (i.e., contain no irrelevant information) and can be parameterized w.r.t. a specification of visible predicates, so that the user may hide uninteresting details from explanations.