6238 papers
This paper empirically demonstrates that while Reinforcement Fine-Tuning (RFT) enables LLM agents to generalize well across varying task difficulties within a single environment, it struggles with cross-environment transfer due to interface and semantic shifts, though sequential and mixture training strategies can effectively mitigate forgetting and improve overall generalization.
Flowcean is a novel, modular framework that automates the data-driven generation and evaluation of Cyber-Physical System models by integrating diverse learning strategies and tools to enhance efficiency and accessibility.
This paper argues that to maintain creative agency while collaborating with emerging intelligent technologies like LLMs, designers must engage in introspection, develop a structural understanding of the technology's capabilities, and deliberately adjust the human-technology working relationship.
This paper presents a systematic sim-to-reality adaptation framework using a high-fidelity Stonefish digital twin and PPO-based Deep Reinforcement Learning to achieve over 90% successful autonomous docking for the Girona AUV, validated by physical experiments demonstrating emergent control behaviors.
This paper introduces LabelPigeon, a novel framework that uses XML tags to jointly perform machine translation and label projection, demonstrating that this approach not only outperforms existing baselines in cross-lingual transfer tasks but also actively improves translation quality across multiple languages.
This paper introduces "Cascade," a framework demonstrating how traditional software and hardware vulnerabilities can be composed with LLM-specific algorithmic weaknesses to amplify adversarial threats and compromise the integrity and confidentiality of compound AI systems.
The paper proposes Slow-Fast Inference (SFI), a training-free framework that accelerates long-context autoregressive decoding by dynamically alternating between low-cost fast steps using stable sparse memory and occasional slow steps that refresh context at semantic boundaries, achieving significant throughput gains without compromising generation quality.
XSkill is a dual-stream framework that enables multimodal agents to continually improve their reasoning and tool orchestration in open-ended settings without parameter updates by extracting and retrieving complementary visual-grounded knowledge in the form of action-level experiences and task-level skills.
This paper proposes a novel training-free method for coarse-guided visual generation that leverages the h-transform to steer diffusion model sampling toward high-fidelity outputs from low-quality references, utilizing a noise-level-aware schedule to balance guidance adherence with synthesis quality without requiring knowledge of the forward transformation operator.
This paper mathematically proves and numerically demonstrates that chemical reaction networks without hidden layers can outperform spiking neural networks with hidden layers in learning classification tasks, offering a theoretical foundation for efficient learning in chemical computers and biological cells.
This paper presents a systematic taxonomy of five families of structured operators that extend or replace standard convolutions in learning-based image processing, providing formal definitions, structural comparisons, and critical analyses of their suitability for various tasks and future research directions.
LoV3D is a novel 3D vision-language pipeline that enhances Alzheimer's disease prognosis by grounding longitudinal MRI analysis in regional volume assessments and a clinically-weighted verifier, achieving state-of-the-art diagnostic accuracy and generalizability while significantly reducing hallucinations through automated, annotation-free training.
This paper proposes a multi-label Temporal Convolutional Network framework that improves transcription factor binding site prediction by modeling cooperative interactions among multiple factors, thereby revealing both known and novel regulatory patterns.
This paper proposes a resource-efficient, closed-loop Neural Architecture Search framework that leverages frozen large language models with a Markov-inspired feedback memory and dual-LLM specialization to iteratively generate and refine compact convolutional neural networks on a single consumer-grade GPU, achieving significant accuracy improvements on image classification benchmarks without requiring cloud infrastructure or model fine-tuning.
This paper introduces LMP2, a browser-based self-audit tool, and reports findings from two user studies demonstrating that LLMs can predict personal features with significant accuracy, while highlighting the operational challenges and nine key frictions in establishing reliable, human-centred privacy audits for generative AI.
This paper proposes a robust Multi-Agent Reinforcement Learning framework for traffic signal control that integrates turning ratio randomization, an exponential phase duration adjustment action space, and a neighbor-based MAPPO observation scheme to significantly reduce average waiting time and improve generalization in dynamic traffic scenarios.
This paper identifies and addresses the "information self-locking" phenomenon in reinforcement learning-trained LLM agents, where deficient action selection and belief tracking create a feedback loop that stifles information gathering, by proposing a method that injects directional critiques to reallocate learning signals and significantly improve active reasoning performance.
This paper introduces Minimax Deep Deterministic Policy Gradient (MMDDPG), a framework that employs a fractional objective to stabilize the minimax optimization between a user policy and an adversarial disturbance policy, thereby learning robust control strategies that maintain performance under external perturbations and model uncertainties in continuous environments.
The paper introduces SommBench, a multilingual benchmark developed in collaboration with professional sommeliers to evaluate the sensory expertise of language models across wine theory, feature completion, and food-wine pairing tasks, revealing that while models excel at theoretical knowledge, they struggle with more complex sensory judgment challenges.
The paper introduces CRAFT, a low-cost, open-source, tendon-driven anthropomorphic hand that combines rigid links with soft joints and rolling-contact surfaces to achieve robust, repeatable, and versatile contact-rich manipulation.