Large Language Models (LLMs) have achieved strong performance on a wide range of complex reasoning tasks, yet further gains are often possible by leveraging the complementary strengths of multiple models. While multi-agent frameworks can improve solution quality by leveraging multiple LLMs, existing methods are often computationally expensive, both at training and inference time. In this work, we introduce a hierarchical multi-agent framework that addresses these challenges by training only a single leader LLM to coordinate a team of untrained peer agents. To this end, we propose Multi-agent guided Leader Policy \textbf{O}ptimization (MLPO), a novel approach which trains the leader to evaluate and synthesize agent responses without auxiliary value networks or explicit agent feedback. Leaders trained with MLPO exhibit improved performance not only when interacting with the agent team at inference time, but also enjoy improved performance when deployed in single-agent settings without the team. Empirical results on Big-Bench Hard (BBH), MATH, and MMLU demonstrate that our framework achieves substantial performance improvements over both single-agent and multi-agent baselines. Our results highlight the effectiveness and efficiency of training a single, flexible leader for collaborative reasoning in multi-agent LLM systems.

Conventional Retrieval Augmented Generation (RAG) approaches are common in text-based applications. However, they struggle with structured, interconnected datasets like knowledge graphs, where understanding underlying relationships is crucial for accurate retrieval. A common direction in graph-based retrieval employs iterative, rule-based traversal guided by Large Language Models (LLMs). Such existing iterative methods typically combine reasoning with single hop traversal at each step, making them vulnerable to LLM reasoning errors and hallucinations that ultimately hinder the retrieval of relevant information. To address these limitations, we propose GraphRunner, a novel graph-based retrieval framework that operates in three distinct stages: planning, verification, and execution. This introduces high-level traversal actions that enable multi-hop exploration in a single step. It also generates a holistic traversal plan, which is verified against the graph structure and pre-defined traversal actions, reducing reasoning errors and detecting hallucinations before execution. GraphRunner significantly reduces LLM reasoning errors and detects hallucinations through validation. Our evaluation using the GRBench dataset shows that GraphRunner consistently outperforms existing approaches, achieving 10-50% performance improvements over the strongest baseline while reducing inference cost by 3.0-12.9x and response generation time by 2.5-7.1x, making it significantly more robust and efficient for graph-based retrieval tasks.

Objectives: To evaluate the current limitations of large language models (LLMs) in medical question answering, focusing on the quality of datasets used for their evaluation. Materials and Methods: Widely-used benchmark datasets, including MedQA, MedMCQA, PubMedQA, and MMLU, were reviewed for their rigor, transparency, and relevance to clinical scenarios. Alternatives, such as challenge questions in medical journals, were also analyzed to identify their potential as unbiased evaluation tools. Results: Most existing datasets lack clinical realism, transparency, and robust validation processes. Publicly available challenge questions offer some benefits but are limited by their small size, narrow scope, and exposure to LLM training. These gaps highlight the need for secure, comprehensive, and representative datasets. Conclusion: A standardized framework is critical for evaluating LLMs in medicine. Collaborative efforts among institutions and policymakers are needed to ensure datasets and methodologies are rigorous, unbiased, and reflective of clinical complexities.

This is the fifth year of the TREC Deep Learning track. As in previous years, we leverage the MS MARCO datasets that made hundreds of thousands of human-annotated training labels available for both passage and document ranking tasks. We mostly repeated last year's design, to get another matching test set, based on the larger, cleaner, less-biased v2 passage and document set, with passage ranking as primary and document ranking as a secondary task (using labels inferred from passage). As we did last year, we sample from MS MARCO queries that were completely held out, unused in corpus construction, unlike the test queries in the first three years. This approach yields a more difficult test with more headroom for improvement. Alongside the usual MS MARCO (human) queries from MS MARCO, this year we generated synthetic queries using a fine-tuned T5 model and using a GPT-4 prompt. The new headline result this year is that runs using Large Language Model (LLM) prompting in some way outperformed runs that use the "nnlm" approach, which was the best approach in the previous four years. Since this is the last year of the track, future iterations of prompt-based ranking can happen in other tracks. Human relevance assessments were applied to all query types, not just human MS MARCO queries. Evaluation using synthetic queries gave similar results to human queries, with system ordering agreement of $τ=0.8487$. However, human effort was needed to select a subset of the synthetic queries that were usable. We did not see clear evidence of bias, where runs using GPT-4 were favored when evaluated using synthetic GPT-4 queries, or where runs using T5 were favored when evaluated on synthetic T5 queries.

The integration of large language model (LLM) technology into judicial systems is fundamentally transforming legal practice worldwide. However, this global transformation has revealed an urgent paradox requiring immediate attention. This study introduces the concept of ``consistency-acceptability divergence'' for the first time, referring to the gap between technical consistency and social acceptance. While LLMs achieve high consistency at the technical level, this consistency demonstrates both positive and negative effects. Through comprehensive analysis of recent data on LLM judicial applications from 2023--2025, this study finds that addressing this challenge requires understanding both task and stakeholder dimensions. This study proposes the Dual-Track Deliberative Multi-Role LLM Judicial Governance Framework (DTDMR-LJGF), which enables intelligent task classification and meaningful interaction among diverse stakeholders. This framework offers both theoretical insights and practical guidance for building an LLM judicial ecosystem that balances technical efficiency with social legitimacy.

Extracting tables and key-value pairs from financial documents is essential for business workflows such as auditing, data analytics, and automated invoice processing. In this work, we introduce Spatial ModernBERT-a transformer-based model augmented with spatial embeddings-to accurately detect and extract tabular data and key-value fields from complex financial documents. We cast the extraction task as token classification across three heads: (1) Label Head, classifying each token as a label (e.g., PO Number, PO Date, Item Description, Quantity, Base Cost, MRP, etc.); (2) Column Head, predicting column indices; (3) Row Head, distinguishing the start of item rows and header rows. The model is pretrained on the PubTables-1M dataset, then fine-tuned on a financial document dataset, achieving robust performance through cross-entropy loss on each classification head. We propose a post-processing method to merge tokens using B-I-IB tagging, reconstruct the tabular layout, and extract key-value pairs. Empirical evaluation shows that Spatial ModernBERT effectively leverages both textual and spatial cues, facilitating highly accurate table and key-value extraction in real-world financial documents.

We propose FLLL3M--Federated Learning with Large Language Models for Mobility Modeling--a privacy-preserving framework for Next-Location Prediction (NxLP). By retaining user data locally and leveraging LLMs through an efficient outer product mechanism, FLLL3M ensures high accuracy with low resource demands. It achieves SOT results on Gowalla (Acc@1: 12.55, MRR: 0.1422), WeePlace (10.71, 0.1285), Brightkite (10.42, 0.1169), and FourSquare (8.71, 0.1023), while reducing parameters by up to 45.6% and memory usage by 52.7%.

Low-Rank Adaptation (LoRA) is one of the most widely used techniques for fine-tuning large language models (LLMs). By introducing a small number of trainable low-rank weight matrices, LoRA substantially reduces the number of parameters that need to be updated, offering significant advantages in memory consumption and computational efficiency compared to full fine-tuning. However, we observed that LoRA does not consistently provide speed improvements across all model architectures and training setups. Motivated by this inconsistency, we conduct a comprehensive analysis of LoRA's performance and investigate the underlying factors limiting its speedup. Based on our findings, we propose several methods for more efficient fine-tuning of LLMs. We empirically evaluate these methods and compare them to LoRA, demonstrating that our approach achieves comparable or superior performance while delivering more consistent training speed improvements. Our work offers valuable insights and practical guidelines for practitioners seeking to optimize LLM fine-tuning under resource constraints.

In an era of rampant misinformation, generating reliable news explanations is vital, especially for under-represented languages like Hindi. Lacking robust automated tools, Hindi faces challenges in scaling misinformation detection. To bridge this gap, we propose a novel framework integrating Direct Preference Optimization (DPO) with curriculum learning to align machine-generated explanations with human reasoning. Fact-checked explanations from credible sources serve as preferred responses, while LLM outputs highlight system limitations and serve as non-preferred responses. To refine task-specific alignment, we introduce two key parameters -- Actuality and Finesse -- into the DPO loss function, enhancing explanation quality and consistency. Experiments with LLMs (Mistral, Llama, Gemma) and PLMs (mBART, mT5) confirm the framework's effectiveness in generating coherent, contextually relevant explanations. This scalable approach combats misinformation and extends automated explanation generation to low-resource languages.

Can machines truly think, reason and act in domains like humans? This enduring question continues to shape the pursuit of Artificial General Intelligence (AGI). Despite the growing capabilities of models such as GPT-4.5, DeepSeek, Claude 3.5 Sonnet, Phi-4, and Grok 3, which exhibit multimodal fluency and partial reasoning, these systems remain fundamentally limited by their reliance on token-level prediction and lack of grounded agency. This paper offers a cross-disciplinary synthesis of AGI development, spanning artificial intelligence, cognitive neuroscience, psychology, generative models, and agent-based systems. We analyze the architectural and cognitive foundations of general intelligence, highlighting the role of modular reasoning, persistent memory, and multi-agent coordination. In particular, we emphasize the rise of Agentic RAG frameworks that combine retrieval, planning, and dynamic tool use to enable more adaptive behavior. We discuss generalization strategies, including information compression, test-time adaptation, and training-free methods, as critical pathways toward flexible, domain-agnostic intelligence. Vision-Language Models (VLMs) are reexamined not just as perception modules but as evolving interfaces for embodied understanding and collaborative task completion. We also argue that true intelligence arises not from scale alone but from the integration of memory and reasoning: an orchestration of modular, interactive, and self-improving components where compression enables adaptive behavior. Drawing on advances in neurosymbolic systems, reinforcement learning, and cognitive scaffolding, we explore how recent architectures begin to bridge the gap between statistical learning and goal-directed cognition. Finally, we identify key scientific, technical, and ethical challenges on the path to AGI.