6987 papers
This paper introduces a novel framework for self-improving LLM agents that automatically extracts structured learnings from execution trajectories—categorizing them into strategy, recovery, and optimization tips—and injects them via adaptive memory retrieval to significantly boost task completion rates, particularly on complex scenarios.
This paper proposes Conditional Expectation Reward (CER), a novel reinforcement learning method that utilizes the large language model itself as an implicit verifier to provide soft, graded reward signals, thereby overcoming the limitations of rule-based verification and enabling effective reasoning training across both mathematical and general free-form answer domains.
This paper proposes a novel, explainable approach to detect and eliminate neural network backdoors by analyzing active paths within the model, demonstrating its effectiveness through experiments on intrusion detection systems.
This paper proposes a novel framework that interleaves task scheduling and motion planning through an incremental learning loop, where symbolic feedback from motion feasibility checks guides the scheduler to generate efficient, collision-free plans for multi-object navigation in shared workspaces.
This paper introduces ROVA, a training framework that enhances video reasoning models' robustness against real-world disturbances like weather and occlusion through difficulty-aware adaptive training and a robustness-aware consistency reward, validated by the new PVRBench benchmark which demonstrates significant performance improvements over existing models.
The paper introduces FAME, a novel abductive explanation method for large neural networks that utilizes dedicated perturbation domains and LiRPA-based bounds to efficiently generate formal abstract minimal explanations, demonstrating superior performance in explanation size and runtime compared to VERIX+.
The paper introduces DxEvolve, a self-evolving diagnostic agent that emulates clinician cognition through an interactive deep clinical research workflow, autonomously requisitioning examinations and externalizing experience to achieve superior diagnostic accuracy and governed continual improvement compared to existing AI models.
This paper presents a platform-agnostic Digital Human Modelling framework that integrates OpenBCI Galea biosensing with a reproducible SuperTux game environment to provide structured, temporally aligned multimodal data for future AI-driven, ethics-approved research in accessibility and inclusive interaction design.
This paper introduces Contract And Conquer (CAC), a black-box adversarial attack method that combines knowledge distillation on an expanding dataset with precise search space contraction to provably compute adversarial examples within a fixed number of iterations, outperforming existing state-of-the-art approaches on ImageNet.
This paper proposes a lightweight, backdoor-based "Intrinsic Proofs" framework that ensures verifiable aggregation in cross-silo federated learning by embedding ephemeral verification signals into model parameters, thereby achieving high detection rates against malicious servers with over 1000x speedup compared to traditional cryptographic methods while preserving client anonymity and final model utility.
This paper proposes RandMark, a random watermarking method that embeds digital watermarks into the internal representations of visual foundation models using a small encoder-decoder network to enable robust ownership verification with low false detection and misdetection rates.
This paper introduces EvoSchema, a comprehensive benchmark featuring a novel taxonomy of ten schema perturbation types to evaluate and enhance the robustness of text-to-SQL models against real-world database schema evolution, revealing that table-level changes significantly impact performance and demonstrating that training on diverse schema designs improves model resilience.
This paper demonstrates that replacing flat text with structured linked data, specifically enhanced entity pages featuring agent instructions and navigational affordances, significantly improves retrieval accuracy and answer quality in both standard and agentic Retrieval-Augmented Generation systems across multiple domains.
AlphaFlowTSE is a one-step conditional generative model for target speaker extraction that utilizes a JVP-free AlphaFlow objective and interval-consistency training to achieve high-fidelity speech recovery with low latency and improved generalization for downstream ASR tasks.
This paper introduces PV-VASM, a model-agnostic probabilistic framework that provides formal robustness guarantees and estimates misclassification probabilities for voice anti-spoofing models against various speech synthesis attacks and unseen perturbations.
This paper proposes CTCNet, a novel cross-spectral guided network featuring a Prototype-Guided Knowledge Embedding module and a Quality-Aware Spectral Compensation module to enhance UAV traffic scene understanding under adverse conditions, accompanied by the introduction of Traffic-VQA, the first large-scale optical-thermal benchmark for cognitive traffic analysis.
This paper introduces HIR-SDD, a novel speech deepfake detection framework that leverages Large Audio Language Models and a human-annotated dataset to achieve robust generalization across audio domains while providing interpretable, human-like reasoning for its predictions.
CUPID is a novel, plug-in framework that enables joint estimation of aleatoric and epistemic uncertainty in pretrained deep learning models without requiring retraining or architectural modifications, thereby enhancing interpretability and trust in high-stakes AI applications.
This paper proposes a deep learning-based autoencoder architecture for the Randomized Distributed Function Computation (RDFC) framework that minimizes the total variation distance to an unknown target distribution using only data samples, demonstrating superior communication efficiency compared to traditional data compression methods, particularly under limited common randomness.
This paper introduces "Super Neurons," a training-free method that leverages scalar activations from the first generated token to create highly accurate classifiers for categorical VQA, enabling extreme early exiting from the first layer and achieving up to a 5.10x speedup while improving performance over the original network.