9743 papers
This paper introduces Direct Consistency Optimization (DCO), a reinforcement learning-inspired method that significantly improves crosslingual knowledge consistency in large language models by deriving a structured reward function directly from the model itself, thereby eliminating the need for an explicit reward model while outperforming existing approaches.
This paper proposes that memory consolidation serves a computational role beyond mere stabilization, utilizing "predictive forgetting" to compress stored representations into a form that optimizes generalization by selectively retaining information that predicts future outcomes, a process necessitated by high-capacity encoding constraints and validated through simulations across diverse neural and transformer models.
This study evaluates the impact of data augmentation and feature enhancement techniques on hate speech detection across traditional and transformer-based models, revealing that while the open-source gpt-oss-20b achieves the highest overall performance, augmentation strategies significantly boost traditional classifiers like Delta TF-IDF and that detection efficacy varies based on the interaction between dataset properties, model architecture, and enhancement methods.
This research demonstrates that fine-tuned Large Language Models, particularly Llama 3.2, significantly outperform traditional machine learning and baseline transformer models in detecting and classifying complex, multilingual illicit content on online marketplaces, offering a scalable solution for enhanced platform safety and law enforcement.
This paper demonstrates that applying the SAM-Audio speech enhancement model as a preprocessing step for zero-shot ASR with Whisper consistently degrades recognition accuracy despite improving perceptual audio quality, revealing a fundamental mismatch between human-perceived signal cleanliness and machine recognition robustness.
This paper introduces "Probabilistic Dreaming," a novel enhancement to the Dreamer world model that utilizes probabilistic methods to enable parallel latent exploration and maintain distinct hypotheses for mutually exclusive futures, resulting in a 4.5% performance improvement and 28% variance reduction on the MPE SimpleTag domain.
This paper proposes a two-layer evaluation framework to assess AI models' ability to simulate justice-specific questioning in moot courts, finding that while models generate realistic questions that cover key legal issues, they still struggle with diversity and sycophancy—shortcomings that naive evaluation methods would miss.
This paper introduces "Model Medicine," a comprehensive clinical research program that treats AI models as biological-like organisms by establishing a taxonomy of subdisciplines, a behavioral genetics framework, and novel diagnostic tools like Neural MRI to systematically understand, diagnose, and treat model disorders.
This paper presents a periodic adaptation framework for unsupervised, pose-based shoplifting detection on edge IoT devices, validated by a new large-scale real-world dataset (RetailS) and demonstrating superior performance and rapid update times compared to offline baselines.
This paper evaluates the viability of zero-shot Multimodal LLMs for real-world video anomaly detection, revealing that while prompt engineering can significantly improve F1-scores, a persistent conservative bias toward the "normal" class severely limits recall, highlighting a critical gap between current MLLM capabilities and the demands of practical surveillance.
This paper presents a neuro-symbolic AI system that autonomously solved an open problem in theoretical physics by deriving novel, exact analytical solutions for the gravitational radiation power spectrum of cosmic strings, thereby surpassing previous partial results and demonstrating AI's potential to accelerate mathematical discovery.
This paper proposes "Interactive Benchmarks," a unified evaluation paradigm that assesses model intelligence through active information acquisition and reasoning under budget constraints in interactive proofs and games, demonstrating that current models still have significant room for improvement in these dynamic scenarios.
This paper challenges the prevailing view of AI memory as a mere functional tool by proposing the "Memory-as-Ontology" paradigm and its associated "Animesis" system, which prioritizes identity continuity and governance over retrieval performance to enable persistent digital citizens that survive model transitions.
This paper introduces CONE, a hybrid transformer encoder that utilizes a novel composite embedding algorithm to preserve the semantics of units and variables for complex numerical data, achieving state-of-the-art performance in numerical reasoning tasks across diverse domains.
The paper introduces DARE, a lightweight retrieval model that integrates data distribution information with function metadata to significantly improve R package retrieval and LLM agent performance in statistical analysis tasks, supported by a new knowledge base and evaluation framework.
This paper argues that the rise of agentic AI necessitates a reconceptualization of Team Situation Awareness to address the structural uncertainty of open-ended agency, distinguishing between continuous alignment mechanisms and emerging tensions to guide future human-AI teaming research.
HiMAP-Travel is a hierarchical multi-agent framework that addresses long-horizon travel planning with hard constraints by employing a Coordinator and parallel Day Executors linked through transactional monitoring and bargaining protocols, achieving state-of-the-art performance and reduced latency on TravelPlanner and FlexTravelBench benchmarks.
To overcome the limitations of existing benchmarks for evaluating Search Agents—such as high construction costs, dynamic obsolescence, attribution ambiguity, and reproducibility issues—this paper introduces Mind-ParaWorld, a novel framework that simulates a parallel world with synthetic, time-shifted scenarios and atomic facts to rigorously assess agents' evidence collection, synthesis, and stopping decisions.
MOOSEnger is a domain-specific AI agent that combines retrieval-augmented generation with deterministic, MOOSE-aware parsing and execution tools to automatically convert natural language into validated simulation inputs, achieving a 93% execution success rate on a diverse benchmark compared to just 8% for an LLM-only baseline.
The paper proposes SharedLLM, a novel framework that extends the context window of large language models to over 128K tokens using a multi-scale self-injection architecture with stacked short-context models and a tree-based retrieval structure, achieving superior performance and efficiency without requiring costly long-context pre-training.