Large language models can generate solutions to complex problems, but training them with reinforcement learning typically requires verifiable rewards that are expensive to create and not possible for all domains. We demonstrate that LLMs can effectively self-improve through self-judging without reference solutions, leveraging the inherent asymmetry between generating and verifying solutions. Our experiments show that models can provide reliable reward signals without ground truth answers, enabling reinforcement learning in domains where verifiable rewards are impractical. By implementing self-judging across Countdown puzzles and integration problems, we achieve performance comparable to formal verification without ground truth solutions. Most notably, Qwen 2.5 7B DeepSeek Distilled trained with self-rewards qualifies for the prestigious MIT Integration Bee competition, performance through self-supervised improvement. When combined with synthetic question generation, we establish a complete self-improvement loop where models generate practice problems, solve them, and evaluate their own performance without any external validation. Our findings demonstrate that LLM judges can provide effective reward signals for training, unlocking reinforcement learning in countless domains previously limited by reward engineering challenges. This work represents a significant step toward autonomous AI systems that continuously improve through self-directed learning rather than human-guided training, potentially accelerating progress across domains where training data is scarce or evaluation is complex.

Visual planning, by offering a sequence of intermediate visual subgoals to a goal-conditioned low-level policy, achieves promising performance on long-horizon manipulation tasks. To obtain the subgoals, existing methods typically resort to video generation models but suffer from model hallucination and computational cost. We present Vis2Plan, an efficient, explainable and white-box visual planning framework powered by symbolic guidance. From raw, unlabeled play data, Vis2Plan harnesses vision foundation models to automatically extract a compact set of task symbols, which allows building a high-level symbolic transition graph for multi-goal, multi-stage planning. At test time, given a desired task goal, our planner conducts planning at the symbolic level and assembles a sequence of physically consistent intermediate sub-goal images grounded by the underlying symbolic representation. Our Vis2Plan outperforms strong diffusion video generation-based visual planners by delivering 53\% higher aggregate success rate in real robot settings while generating visual plans 35$\times$ faster. The results indicate that Vis2Plan is able to generate physically consistent image goals while offering fully inspectable reasoning steps.

This research prepares an automatic pipeline for generating reliable question-answer (Q&A) tests using AI chatbots. We automatically generated a GPT-4o-mini-based Q&A test for a Natural Language Processing course and evaluated its psychometric and perceived-quality metrics with students and experts. A mixed-format IRT analysis showed that the generated items exhibit strong discrimination and appropriate difficulty, while student and expert star ratings reflect high overall quality. A uniform DIF check identified two items for review. These findings demonstrate that LLM-generated assessments can match human-authored tests in psychometric performance and user satisfaction, illustrating a scalable approach to AI-assisted assessment development.

Can LLMs Be Trusted for Evaluating RAG Systems? Abstract--Retrieval-Augmented Generation (RAG) has advanced significantly in recent years. The complexity of RAG systems, which involve multiple components--such as indexi ng, retrieval, and generation--along with numerous other param e-ters, poses substantial challenges for systematic evaluat ion and quality enhancement. Previous research highlights that ev aluating RAG systems is essential for documenting advancements, com - paring configurations, and identifying effective approach es for domain-specific applications. This study systematically r eviews 63 academic articles to provide a comprehensive overview of state-of-the-art RAG evaluation methodologies, focusing on four key areas: datasets, retrievers, indexing and databases, a nd the generator component. We observe the feasibility of an automated evaluation approach for each component of a RAG system, leveraging an LLM capable of both generating evalua tion datasets and conducting evaluations. In addition, we found that further practical research is essential to provide compani es with clear guidance on the do's and don'ts of implementing and evaluating RAG systems. By synthesizing evaluation approa ches for key RAG components and emphasizing the creation and adaptation of domain-specific datasets for benchmarking, w e contribute to the advancement of systematic evaluation met hods and the improvement of evaluation rigor for RAG systems. Furthermore, by examining the interplay between automated approaches leveraging LLMs and human judgment, we contribute to the ongoing discourse on balancing automation and human input, clarifying their respective contributions, limita tions, and challenges in achieving robust and reliable evaluations. In recent years, Large Language Models (LLMs) have made significant progress in research and have grown increasingl y popular [1]. However, LLMs face several challenges, includ - ing issues with hallucinations caused by insufficient conte xt [2], as well as limitations in their learned content, which prevent them from addressing questions requiring specific or proprietary information [1].

Large language models (LLMs) demonstrate strong capabilities in reasoning and question answering, yet their tendency to generate factually incorrect content remains a critical challenge. This study evaluates proprietary and open-source LLMs on generating relevant research papers with accurate arXiv links. Our evaluation reveals critical academic risks: LLMs frequently generate incorrect arXiv links or references to non-existent papers, fundamentally undermining their ability to properly attribute research contributions to the actual authors. We introduce arXivBench, a benchmark specifically designed to assess LLM performance across eight major subject categories on arXiv and five subfields within computer science, one of the most popular categories among them. Our findings show concerning accuracy variations across subjects, with Claude-3.5-Sonnet exhibiting a substantial advantage in generating both relevant and accurate responses. Notably, most LLMs perform significantly better in Artificial Intelligence than other subfields. This benchmark provides a standardized tool for evaluating LLM reliability in scientific contexts, promoting more dependable academic use in research environments. Our code and dataset are available at https://github.com/liningresearch/arXivBench and https://huggingface.co/datasets/arXivBenchLLM/arXivBench.

Conversational Recommender Systems (CRSs)aim to engage users in dialogue to provide tailored recommendations. While traditional CRSs focus on eliciting preferences and retrieving items, real-world e-commerce interactions involve more complex decision-making, where users consider multiple factors beyond simple attributes. To capture this complexity, we introduce Conversational Sales (CSALES), a novel task that integrates preference elicitation, recommendation, and persuasion within a unified conversational framework. To support realistic and systematic evaluation, we present CSUSER, an evaluation protocol with LLM-based user simulator grounded in real-world behavioral data by modeling fine-grained user profiles for personalized interaction. We also propose CSI, a conversational sales agent that proactively infers contextual user profiles and strategically selects actions through conversation. Comprehensive experiments show that CSI significantly improves both recommendation success and persuasive effectiveness across diverse user profiles.

Speculative Decoding (SD) is a widely used approach to accelerate the inference of large language models (LLMs) without reducing generation quality. It operates by first using a compact model to draft multiple tokens efficiently, followed by parallel verification using the target LLM. This approach leads to faster inference compared to auto-regressive decoding. While there are multiple approaches to create a draft model, one promising approach is to use early-exit methods. These methods draft candidate tokens by using a subset of layers of the primary model and applying the remaining layers for verification, allowing a single model to handle both drafting and verification. While this technique reduces memory usage and computational cost, its performance relies on the choice of the exit layer for drafting and the number of tokens drafted (speculation length) in each SD round. Prior works use hyperparameter exploration to statically select these values. However, our evaluations show that these hyperparameter values are task-specific, and even within a task they are dependent on the current sequence context. We introduce DEL (Dynamic Exit Layer), a plug-and-play method that adaptively selects the exit layer and speculation length during inference. DEL dynamically tracks the token acceptance rate if the tokens are drafted at each layer of an LLM and uses that knowledge to heuristically select the optimal exit layer and speculation length. Our experiments across a broad range of models and downstream tasks show that DEL achieves overall speedups of $2.16\times$$\sim$$2.62\times$ over vanilla auto-regressive decoding and improves upon state-of-the-art SD methods, which peak at $2.43\times$, by up to $0.19\times$. The code is available at https://github.com/hoenza/DEL.

Truly multilingual safety moderation efforts for Large Language Models (LLMs) have been hindered by a narrow focus on a small set of languages (e.g., English, Chinese) as well as a limited scope of safety definition, resulting in significant gaps in moderation capabilities. To bridge these gaps, we release POLYGUARD, a new state-of-the-art multilingual safety model for safeguarding LLM generations, and the corresponding training and evaluation datasets. POLYGUARD is trained on POLYGUARDMIX, the largest multilingual safety training corpus to date containing 1.91M samples across 17 languages (e.g., Chinese, Czech, English, Hindi). We also introduce POLYGUARDPROMPTS, a high quality multilingual benchmark with 29K samples for the evaluation of safety guardrails. Created by combining naturally occurring multilingual human-LLM interactions and human-verified machine translations of an English-only safety dataset (WildGuardMix; Han et al., 2024), our datasets contain prompt-output pairs with labels of prompt harmfulness, response harmfulness, and response refusal. Through extensive evaluations across multiple safety and toxicity benchmarks, we demonstrate that POLYGUARD outperforms existing state-of-the-art open-weight and commercial safety classifiers by 5.5%. Our contributions advance efforts toward safer multilingual LLMs for all global users.

Traditional control and planning for robotic manipulation heavily rely on precise physical models and predefined action sequences. While effective in structured environments, such approaches often fail in real-world scenarios due to modeling inaccuracies and struggle to generalize to novel tasks. In contrast, humans intuitively interact with their surroundings, demonstrating remarkable adaptability, making efficient decisions through implicit physical understanding. In this work, we propose INTENTION, a novel framework enabling robots with learned interactive intuition and autonomous manipulation in diverse scenarios, by integrating Vision-Language Models (VLMs) based scene reasoning with interaction-driven memory. We introduce Memory Graph to record scenes from previous task interactions which embodies human-like understanding and decision-making about different tasks in real world. Meanwhile, we design an Intuitive Perceptor that extracts physical relations and affordances from visual scenes. Together, these components empower robots to infer appropriate interaction behaviors in new scenes without relying on repetitive instructions. Videos: https://robo-intention.github.io

Large Language Models (LLMs) are increasingly integrated with graph-structured data for tasks like node classification, a domain traditionally dominated by Graph Neural Networks (GNNs). While this integration leverages rich relational information to improve task performance, their robustness against adversarial attacks remains unexplored. We take the first step to explore the vulnerabilities of graph-aware LLMs by leveraging existing adversarial attack methods tailored for graph-based models, including those for poisoning (training-time attacks) and evasion (test-time attacks), on two representative models, LLAGA (Chen et al. 2024) and GRAPHPROMPTER (Liu et al. 2024). Additionally, we discover a new attack surface for LLAGA where an attacker can inject malicious nodes as placeholders into the node sequence template to severely degrade its performance. Our systematic analysis reveals that certain design choices in graph encoding can enhance attack success, with specific findings that: (1) the node sequence template in LLAGA increases its vulnerability; (2) the GNN encoder used in GRAPHPROMPTER demonstrates greater robustness; and (3) both approaches remain susceptible to imperceptible feature perturbation attacks. Finally, we propose an end-to-end defense framework GALGUARD, that combines an LLM-based feature correction module to mitigate feature-level perturbations and adapted GNN defenses to protect against structural attacks.