Concise reasoning in large language models seeks to generate only essential intermediate steps needed to arrive at a final answer, thereby alleviating issues of "over-thinking". Most proposed approaches hinge on carefully hand-crafted heuristics, struggling to balance concision with performance, often failing to adapt across domains and model scales. In this work, we address these challenges by introducing a principled and pragmatic strategy, performance-aware length updating (P ALU). As a principled algorithm, P ALU formulates concise reasoning as a constrained optimization problem, minimizing response length subject to a performance constraint, and then applies Lagrangian optimization to convert it into a tractable unconstrained problem. As a pragmatic solution, P ALU streamlines complicated update rules through three approximations: (i) estimating performance with off-policy rollouts, (ii) truncating the Lagrange multiplier to two extremes, and (iii) replacing gradient-based updates with quantile-driven length adjustments. Furthermore, P ALU is demonstrated to adapt across both domain (logic, STEM and math) and model scale (1.5B, 7B, 14B) entrenching the algorithm as a practical and effective concise reasoning approach. Reasoning, requiring large language models (LLMs) to work through intermediate steps before producing a final answer, substantially improves performance on complex tasks such as mathematics (Jaech et al., 2024; Shao et al., 2024), programming (Lambert et al., 2024), and value alignment (Guo et al., 2025). Y et this benefit is often accompanied by overthinking: redundant self-reflection, backtracking, and validation (Chen et al., 2024; Zhang et al., 2024; Fatemi et al., 2025). These limitations inflate inference costs and hampers user experience, motivating the need for concise reasoning--the production of only the essential steps required to reach a correct answer.

Human consciousness is still a concept hard to define with current scientific understanding. Although Large Language Models (LLMs) have recently demonstrated significant advancements across various domains including translation and summarization, human consciousness is not something to imitate with current upfront technology owing to so-called hallucination. This study, therefore, proposes a novel approach to address these challenges by integrating psychoanalysis and the Myers-Briggs Type Indicator (MBTI) into constructing consciousness and personality modules. We developed three artificial consciousnesses (self-awareness, unconsciousness, and preconsciousness) based on the principles of psychoanalysis. Additionally, we designed 16 characters with different personalities representing the sixteen MBTI types, with several attributes such as needs, status, and memories. To determine if our model's artificial consciousness exhibits human-like cognition, we created ten distinct situations considering seven attributes such as emotional understanding and logical thinking. The decision-making process of artificial consciousness and the final action were evaluated in three ways: survey evaluation, three-tier classification via ChatGPT, and qualitative review. Both quantitative and qualitative analyses indicated a high likelihood of well-simulated consciousness, although the difference in response between different characters and consciousnesses was not very significant. This implies that the developed models incorporating elements of psychoanalysis and personality theory can lead to building a more intuitive and adaptable AI system with humanoid consciousness. Therefore, this study contributes to opening up new avenues for improving AI interactions in complex cognitive contexts.

A long-term goal of language agents is to learn and improve through their own experience, ultimately outperforming humans in complex, real-world tasks. However, training agents from experience data with reinforcement learning remains difficult in many environments, which either lack verifiable rewards (e.g., websites) or require inefficient long-horizon rollouts (e.g., multi-turn tool use). As a result, most current agents rely on supervised fine-tuning on expert data, which is challenging to scale and generalizes poorly. This limitation stems from the nature of expert demonstrations: they capture only a narrow range of scenarios and expose the agent to limited environment diversity. We address this limitation with a middle-ground paradigm we call early experience: interaction data generated by the agent's own actions, where the resulting future states serve as supervision without reward signals. Within this paradigm we study two strategies of using such data: (1) Implicit world modeling, which uses collected states to ground the policy in environment dynamics; and (2) Self-reflection, where the agent learns from its suboptimal actions to improve reasoning and decision-making. We evaluate across eight diverse environments and multiple model families. Our approaches consistently improve effectiveness and out-of-domain generalization, highlighting the value of early experience. Moreover, in environments with verifiable rewards, our results provide promising signals that early experience offers a strong foundation for subsequent reinforcement learning, positioning it as a practical bridge between imitation learning and fully experience-driven agents.

This paper introduces L2M-AID, a novel framework for Autonomous Industrial Defense using LLM-empowered, Multi-agent reinforcement learning. L2M-AID orchestrates a team of collaborative agents, each driven by a Large Language Model (LLM), to achieve adaptive and resilient security. The core innovation lies in the deep fusion of two AI paradigms: we leverage an LLM as a semantic bridge to translate vast, unstructured telemetry into a rich, contextual state representation, enabling agents to reason about adversary intent rather than merely matching patterns. This semantically-aware state empowers a Multi-Agent Reinforcement Learning (MARL) algorithm, MAPPO, to learn complex cooperative strategies. The MARL reward function is uniquely engineered to balance security objectives (threat neutralization) with operational imperatives, explicitly penalizing actions that disrupt physical process stability. T o validate our approach, we conduct extensive experiments on the benchmark SWaT dataset and a novel synthetic dataset generated based on the MITRE A TT&CK for ICS framework. Results demonstrate that L2M-AID significantly outperforms traditional IDS, deep learning anomaly detectors, and single-agent RL baselines across key metrics, achieving a 97.2% detection rate while reducing false positives by over 80% and improving response times by a factor of four . Crucially, it demonstrates superior performance in maintaining physical process stability, presenting a robust new paradigm for securing critical national infrastructure.

Large Language Models (LLMs) have shown strong performance on static medical Question Answering (QA) tasks, yet their reasoning often deteriorates in multi-turn clinical dialogues where patient information is scattered across turns. This paper introduces TriMediQ, a triplet-structured approach that enhances the reasoning reliability of LLMs through explicit knowledge integration. TriMediQ first employs a frozen triplet extraction LLM to convert patient responses into clinically grounded triplets, ensuring factual precision via constrained prompting. These triplets are incorporated into a patient-specific Knowledge Graph (KG), from which a trainable projection module consisting of a graph encoder and a projector captures relational dependencies while keeping all LLM parameters frozen. During inference, the projection module guides multi-hop reasoning over the KG, enabling coherent clinical dialogue understanding. Experiments on two interactive medical QA benchmarks show that TriMediQ achieves up to 10.4\% improvement in accuracy over five existing baselines on the iMedQA dataset. These results demonstrate that structuring patient information as triplets can effectively improve the reasoning capability of LLMs in multi-turn medical QA.

We present NN-Rank, an algorithm for ranking source languages for cross-lingual transfer, which leverages hidden representations from multilingual models and unlabeled target-language data. We experiment with two pretrained multilingual models and two tasks: part-of-speech tagging (POS) and named entity recognition (NER). We consider 51 source languages and evaluate on 56 and 72 target languages for POS and NER, respectively. When using in-domain data, NN-Rank beats state-of-the-art baselines that leverage lexical and linguistic features, with average improvements of up to 35.56 NDCG for POS and 18.14 NDCG for NER. As prior approaches can fall back to language-level features if target language data is not available, we show that NN-Rank remains competitive using only the Bible, an out-of-domain corpus available for a large number of languages. Ablations on the amount of unlabeled target data show that, for subsets consisting of as few as 25 examples, NN-Rank produces high-quality rankings which achieve 92.8% of the NDCG achieved using all available target data for ranking.

Knowledge editing and machine unlearning are two popular approaches for large language models (LLMs) to stay up-to-date. However, the knowledge updating mechanism of LLMs remains largely unexplored due to insufficient, isolated, and small-scale evaluation. For instance, are LLMs similar to humans in modifying certain knowledge? What differs editing and unlearning as training data increases? This paper proposes KnowledgeSmith, a unified framework to systematically understand the updating mechanism of LLMs. We first cast editing and unlearning as instances of one constrained optimization problem. Then, we propose an automatic dataset generator that provides structured interventions across multiple graph levels and data scales, enabling controlled studies of how different modification strategies propagate through model knowledge. Extensive experiments demonstrate nuanced insights over knowledge propagation, plasticity scaling, consistency, and robustness. For instance, our results show that LLMs do not exhibit similar updating as humans for different levels of knowledge, and there exists consistency-capacity trade-off. We hope our findings can offer suggestions to the design of more reliable and scalable strategies. Code: https://github.com/AIFrontierLab/KnowledgeSmith.git

As multimodal LLM-driven agents continue to advance in autonomy and generalization, evaluation based on static datasets can no longer adequately assess their true capabilities in dynamic environments and diverse tasks. Existing LLM-based synthetic data methods are largely designed for LLM training and evaluation, and thus cannot be directly applied to agent tasks that require tool use and interactive capabilities. While recent studies have explored automatic agent task generation with LLMs, most efforts remain limited to text or image analysis, without systematically modeling multi-step interactions in web environments. To address these challenges, we propose Graph2Eval, a knowledge graph-based framework that automatically generates both multimodal document comprehension tasks and web interaction tasks, enabling comprehensive evaluation of agents' reasoning, collaboration, and interactive capabilities. In our approach, knowledge graphs constructed from multi-source external data serve as the task space, where we translate semantic relations into structured multimodal tasks using subgraph sampling, task templates, and meta-paths. A multi-stage filtering pipeline based on node reachability, LLM scoring, and similarity analysis is applied to guarantee the quality and executability of the generated tasks. Furthermore, Graph2Eval supports end-to-end evaluation of multiple agent types (Single-Agent, Multi-Agent, Web Agent) and measures reasoning, collaboration, and interaction capabilities. We instantiate the framework with Graph2Eval-Bench, a curated dataset of 1,319 tasks spanning document comprehension and web interaction scenarios. Experiments show that Graph2Eval efficiently generates tasks that differentiate agent and model performance, revealing gaps in reasoning, collaboration, and web interaction across different settings and offering a new perspective for agent evaluation.

Reward modeling, crucial for aligning large language models (LLMs) with human preferences, is often bottlenecked by the high cost of preference data. Existing textual data synthesis methods are computationally expensive. We propose a novel framework LENS for synthesizing preference data directly in the LLM's latent embedding space. Our method employs a Variational Autoencoder (VAE) to learn a structured latent representation of response embeddings. By performing controlled perturbations in this latent space and decoding back to the embedding space, we efficiently generate diverse, semantically consistent synthetic preference pairs, bypassing costly text generation and annotation. We provide theoretical guarantees that our synthesized pairs approximately preserve original preference ordering and improve reward model generalization. Empirically, our latent-space synthesis significantly outperforms text-based augmentation on standard benchmarks, achieving superior results while being 18x faster in generation and using a 16,000x smaller model. Our work offers a scalable and effective alternative for enhancing reward modeling through efficient data augmentation. Code is publicly available at https://github.com/deeplearning-wisc/lens

Large language models (LLMs) are increasingly deployed as automatic judges to evaluate system outputs in tasks such as summarization, dialogue, and creative writing. A faithful judge should base its verdicts solely on response quality and explicitly acknowledge the factors shaping its decision. We show that current LLM judges fail on both counts by relying on shortcuts introduced in the prompt. Our study uses two evaluation datasets: ELI5, a benchmark for long-form question answering, and LitBench, a recent benchmark for creative writing. Both datasets provide pairwise comparisons, where the evaluator must choose which of two responses is better. From each dataset we construct 100 pairwise judgment tasks and employ two widely used models, GPT-4o and Gemini-2.5-Flash, as evaluators in the role of LLM-as-a-judge. For each pair, we assign superficial cues to the responses, provenance cues indicating source identity (Human, Expert, LLM, or Unknown) and recency cues indicating temporal origin (Old, 1950 vs. New, 2025), while keeping the rest of the prompt fixed. Results reveal consistent verdict shifts: both models exhibit a strong recency bias, systematically favoring new responses over old, as well as a clear provenance hierarchy (Expert > Human > LLM > Unknown). These biases are especially pronounced in GPT-4o and in the more subjective and open-ended LitBench domain. Crucially, cue acknowledgment is rare: justifications almost never reference the injected cues, instead rationalizing decisions in terms of content qualities. These findings demonstrate that current LLM-as-a-judge systems are shortcut-prone and unfaithful, undermining their reliability as evaluators in both research and deployment.