The rapid evolution of large language models (LLMs) has revolutionized various fields, including the identification and discovery of human values within text data. While traditional NLP models, such as BERT, have been employed for this task, their ability to represent textual data is significantly outperformed by emerging LLMs like GPTs. However, the performance of online LLMs often degrades when handling long contexts required for value identification, which also incurs substantial computational costs. To address these challenges, we propose EAVIT, an efficient and accurate framework for human value identification that combines the strengths of both locally fine-tunable and online black-box LLMs. Our framework employs a value detector - a small, local language model - to generate initial value estimations. These estimations are then used to construct concise input prompts for online LLMs, enabling accurate final value identification. To train the value detector, we introduce explanation-based training and data generation techniques specifically tailored for value identification, alongside sampling strategies to optimize the brevity of LLM input prompts. Our approach effectively reduces the number of input tokens by up to 1/6 compared to directly querying online LLMs, while consistently outperforming traditional NLP methods and other LLM-based strategies.

In recent years, large language models (LLMs) have demonstrated remarkable generalization capabilities across various natural language processing (NLP) tasks. Similarly, graph foundation models (GFMs) have emerged as a promising direction in graph learning, aiming to generalize across diverse datasets through large-scale pre-training. However, unlike language models that rely on explicit token representations, graphs lack a well-defined unit for generalization, making it challenging to design effective pre-training strategies. In this work, we propose REEF, a novel framework that leverages relation tokens as the basic units for GFMs. Inspired by the token vocabulary in LLMs, we construct a relation vocabulary of relation tokens to store relational information within graphs. To accommodate diverse relations, we introduce two hypernetworks that adaptively generate the parameters of aggregators and classifiers in graph neural networks based on relation tokens. In addition, we design another hypernetwork to construct dataset-specific projectors and incorporate a dataset-level feature bias into the initial node representations, enhancing flexibility across different datasets with the same relation. Further, we adopt graph data augmentation and a mixed-dataset pre-training strategy, allowing REEF to capture relational diversity more effectively and exhibit strong generalization capabilities. Extensive experiments show that REEF significantly outperforms existing methods on both pre-training and transfer learning tasks, underscoring its potential as a powerful foundation model for graph-based applications.

Large Language Model (LLM) reasoning for complex tasks inherently involves a trade-off between solution accuracy and computational efficiency. The subsequent step of verification, while intended to improve performance, further complicates this landscape by introducing its own challenging trade-off: sophisticated Generative Reward Models (GenRMs) can be computationally prohibitive if naively integrated with LLMs at test-time, while simpler, faster methods may lack reliability. To overcome these challenges, we introduce FlexiVe, a novel generative verifier that flexibly balances computational resources between rapid, reliable fast thinking and meticulous slow thinking using a Flexible Allocation of Verification Budget strategy. We further propose the Solve-Detect-Verify pipeline, an efficient inference-time scaling framework that intelligently integrates FlexiVe, proactively identifying solution completion points to trigger targeted verification and provide focused solver feedback. Experiments show FlexiVe achieves superior accuracy in pinpointing errors within reasoning traces on ProcessBench. Furthermore, on challenging mathematical reasoning benchmarks (AIME 2024, AIME 2025, and CNMO), our full approach outperforms baselines like self-consistency in reasoning accuracy and inference efficiency. Our system offers a scalable and effective solution to enhance LLM reasoning at test time.

We present Team asdfo123's submission to the LLMSR@XLLM25 shared task, which evaluates large language models on producing fine-grained, controllable, and interpretable reasoning processes. Systems must extract all problem conditions, decompose a chain of thought into statement-evidence pairs, and verify the logical validity of each pair. Leveraging only the off-the-shelf Meta-Llama-3-8B-Instruct, we craft a concise few-shot, multi-turn prompt that first enumerates all conditions and then guides the model to label, cite, and adjudicate every reasoning step. A lightweight post-processor based on regular expressions normalises spans and enforces the official JSON schema. Without fine-tuning, external retrieval, or ensembling, our method ranks 5th overall, achieving macro F1 scores on par with substantially more complex and resource-consuming pipelines. We conclude by analysing the strengths and limitations of our approach and outlining directions for future research in structural reasoning with LLMs. Our code is available at https://github.com/asdfo123/LLMSR-asdfo123.

With the proliferation of large language models (LLMs) in the medical domain, there is increasing demand for improved evaluation techniques to assess their capabilities. However, traditional metrics like F1 and ROUGE, which rely on token overlaps to measure quality, significantly overlook the importance of medical terminology. While human evaluation tends to be more reliable, it can be very costly and may as well suffer from inaccuracies due to limits in human expertise and motivation. Although there are some evaluation methods based on LLMs, their usability in the medical field is limited due to their proprietary nature or lack of expertise. To tackle these challenges, we present AutoMedEval, an open-sourced automatic evaluation model with 13B parameters specifically engineered to measure the question-answering proficiency of medical LLMs. The overarching objective of AutoMedEval is to assess the quality of responses produced by diverse models, aspiring to significantly reduce the dependence on human evaluation. Specifically, we propose a hierarchical training method involving curriculum instruction tuning and an iterative knowledge introspection mechanism, enabling AutoMedEval to acquire professional medical assessment capabilities with limited instructional data. Human evaluations indicate that AutoMedEval surpasses other baselines in terms of correlation with human judgments.

Augmenting large language models (LLMs) with auxiliary tokens has emerged as a promising strategy for enhancing model performance. In this work, we introduce a lightweight method termed latent tokens; these are dummy tokens that may be non-interpretable in natural language but steer the autoregressive decoding process of a Transformer-based LLM via the attention mechanism. The proposed latent tokens can be seamlessly integrated with a pre-trained Transformer, trained in a parameter-efficient manner, and applied flexibly at inference time, while adding minimal complexity overhead to the existing infrastructure of standard Transformers. We propose several hypotheses about the underlying mechanisms of latent tokens and design synthetic tasks accordingly to verify them. Numerical results confirm that the proposed method noticeably outperforms the baselines, particularly in the out-of-distribution generalization scenarios, highlighting its potential in improving the adaptability of LLMs.

Hallucination, the generation of factually incorrect information, remains a significant challenge for large language models (LLMs), especially in open-domain long-form generation. Existing approaches for detecting hallucination in long-form tasks either focus on limited domains or rely heavily on external fact-checking tools, which may not always be available. In this work, we systematically investigate reference-free hallucination detection in open-domain long-form responses. Our findings reveal that internal states (e.g., model's output probability and entropy) alone are insufficient for reliably (i.e., better than random guessing) distinguishing between factual and hallucinated content. To enhance detection, we explore various existing approaches, including prompting-based methods, probing, and fine-tuning, with fine-tuning proving the most effective. To further improve the accuracy, we introduce a new paradigm, named RATE-FT, that augments fine-tuning with an auxiliary task for the model to jointly learn with the main task of hallucination detection. With extensive experiments and analysis using a variety of model families & datasets, we demonstrate the effectiveness and generalizability of our method, e.g., +3% over general fine-tuning methods on LongFact.

Out-of-distribution (OOD) detection is critical for ensuring the safety and reliability of machine learning systems, particularly in dynamic and open-world environments. In the vision and text domains, zero-shot OOD detection - which requires no training on in-distribution (ID) data - has advanced significantly through the use of large-scale pretrained models, such as vision-language models (VLMs) and large language models (LLMs). However, zero-shot OOD detection in graph-structured data remains largely unexplored, primarily due to the challenges posed by complex relational structures and the absence of powerful, large-scale pretrained models for graphs. In this work, we take the first step toward enabling zero-shot graph OOD detection by leveraging a graph foundation model (GFM). Our experiments show that, when provided only with class label names for both ID and OOD categories, the GFM can effectively perform OOD detection - often surpassing existing "supervised" OOD detection methods that rely on extensive labeled node data. We further address the practical scenario in which OOD label names are not available in real-world settings by introducing GLIP-OOD, a framework that uses LLMs to generate semantically informative pseudo-OOD labels from unlabeled data. These generated OOD labels allow the GFM to better separate ID and OOD classes, facilitating more precise OOD detection - all without any labeled nodes (only ID label names). To our knowledge, this is the first approach to achieve node-level graph OOD detection in a fully zero-shot setting, and it attains performance comparable to state-of-the-art supervised methods on four benchmark text-attributed graph datasets.

AI-driven models have shown great promise in detecting errors in radiology reports, yet the field lacks a unified benchmark for rigorous evaluation of error detection and further correction. To address this gap, we introduce CorBenchX, a comprehensive suite for automated error detection and correction in chest X-ray reports, designed to advance AI-assisted quality control in clinical practice. We first synthesize a large-scale dataset of 26,326 chest X-ray error reports by injecting clinically common errors via prompting DeepSeek-R1, with each corrupted report paired with its original text, error type, and human-readable description. Leveraging this dataset, we benchmark both open- and closed-source vision-language models,(e.g., InternVL, Qwen-VL, GPT-4o, o4-mini, and Claude-3.7) for error detection and correction under zero-shot prompting. Among these models, o4-mini achieves the best performance, with 50.6 % detection accuracy and correction scores of BLEU 0.853, ROUGE 0.924, BERTScore 0.981, SembScore 0.865, and CheXbertF1 0.954, remaining below clinical-level accuracy, highlighting the challenge of precise report correction. To advance the state of the art, we propose a multi-step reinforcement learning (MSRL) framework that optimizes a multi-objective reward combining format compliance, error-type accuracy, and BLEU similarity. We apply MSRL to QwenVL2.5-7B, the top open-source model in our benchmark, achieving an improvement of 38.3% in single-error detection precision and 5.2% in single-error correction over the zero-shot baseline.

Instruction data selection aims to identify a high-quality subset from the training set that matches or exceeds the performance of the full dataset on target tasks. Existing methods focus on the instruction-to-response mapping, but neglect the human preference for diverse responses. In this paper, we propose Preference-oriented Data Selection method (ProDS) that scores training samples based on their alignment with preferences observed in the target set. Our key innovation lies in shifting the data selection criteria from merely estimating features for accurate response generation to explicitly aligning training samples with human preferences in target tasks. Specifically, direct preference optimization (DPO) is employed to estimate human preferences across diverse responses. Besides, a bidirectional preference synthesis strategy is designed to score training samples according to both positive preferences and negative preferences. Extensive experimental results demonstrate our superiority to existing task-agnostic and targeted methods.