Recent advancements have underscored the potential of large language model (LLM)-based agents in financial decision-making. Despite this progress, the field currently encounters two main challenges: (1) the lack of a comprehensive LLM agent framework adaptable to a variety of financial tasks, and (2) the absence of standardized benchmarks and consistent datasets for assessing agent performance. To tackle these issues, we introduce InvestorBench, the first benchmark specifically designed for evaluating LLM-based agents in diverse financial decision-making contexts. InvestorBench enhances the versatility of LLM-enabled agents by providing a comprehensive suite of tasks applicable to different financial products, including single equities like stocks and cryptocurrencies, and exchange-traded funds (ETFs). Additionally, we assess the reasoning and decision-making capabilities of our agent framework using thirteen different LLMs as backbone models, across various market environments and tasks. Furthermore, we have curated a diverse collection of open-source, datasets and developed a comprehensive suite of environments for financial decision-making. This establishes a highly accessible platform for evaluating financial agents’ performance across various scenarios.

Large language models (LLMs) fine-tuned on multimodal financial data have demonstrated impressive reasoning capabilities in various financial tasks. However, they often struggle with multi-step, goal-oriented scenarios in interactive financial markets, such as trading, where complex agentic approaches are required to improve decision-making. To address this, we propose FLAG-Trader, a unified architecture integrating linguistic processing (via LLMs) with gradient-driven reinforcement learning (RL) policy optimization, in which a partially fine-tuned LLM acts as the policy network, leveraging pre-trained knowledge while adapting to the financial domain through parameter-efficient fine-tuning. Through policy gradient optimization driven by trading rewards, our framework not only enhances LLM performance in trading but also improves results on other financial-domain tasks. We present extensive empirical evidence to validate these enhancements.

Recent advances in AI-generated video have shown strong performance on text-to-video tasks, particularly for short clips depicting a single scene. However, current models struggle to generate longer videos with coherent scene transitions, primarily because they cannot infer when a transition is needed from the prompt. Most open-source models are trained on datasets consisting of single-scene video clips, which limits their capacity to learn and respond to prompts requiring multiple scenes. Developing scene transition awareness is essential for multi-scene generation, as it allows models to identify and segment videos into distinct clips by accurately detecting transitions. To address this, we introduce the Transition-Aware Video (TAV) dataset with multi-scene clips and captions that explicitly state scene segmentation and transition structure. Our focus is on how prompt semantics and dataset annotations about temporal context affect text-to-video generation. Post-training on TAV improves alignment between the scene count implied by prompt and the scene count produced by the model, while preserving visual quality.

Financial misinformation poses a significant threat to investment decisions and market stability. Recently, the application of Large Language Models (LLMs) for detecting financial misinformation has gained considerable attention within the natural language processing (NLP) community. The Financial Misinformation Detection (FMD) challenge @ Coling 2025 serves as a valuable platform for collaboration and innovation. This paper presents our solution to FMD challenge. Our approach involves using search engines to retrieve the summarized high-quality information as supporting evidence and designing a financial domain-specific chain-of-thought to enhance the reasoning capabilities of LLMs. We evaluated our method on both commercial closed-source LLMs (GPT-family) and open-source models (Llama-3.1-8B and QWen). The experimental results domonstrate that the proposed method improves veracity prediction performance. However, the quality of the generated explanations remains relatively poor. In the paper, we present the experimental findings and provides an in depth analysis of these results.

Despite the promise of large language models based agent framework in stock trading task, their capabilities for comprehensive analysis and multiple different financial assets remain largely unexplored, such as cryptocurrency trading. To evaluate the capabilities of LLM-based agent framework in cryptocurrency trading, we introduce an LLMs-based financial shared task featured at COLING 2025 FinNLP-FNP-LLMFinLegal workshop, named Agent-based Single Cryptocurrency Trading Challenge. This challenge includes two cryptocurrencies: BitCoin and Ethereum. In this paper, we provide an overview of these tasks and datasets, summarize participants’ methods, and present their experimental evaluations, highlighting the effectiveness of LLMs in addressing cryptocurrency trading challenges. To the best of our knowledge, the Agent-based Single Cryptocurrency Trading Challenge is one of the first challenges for assessing LLMs in the financial area. In consequence, we provide detailed observations and take away conclusions for future development in this area.

Large language models (LLMs) are increasingly being used for complex research tasks such as literature review, idea generation, and scientific paper analysis, yet their ability to truly understand and process the intricate relationships within complex research papers, such as the logical links between claims and supporting evidence remains largely unexplored. In this study, we present CLAIM-BENCH, a comprehensive benchmark for evaluating LLMs’ capabilities in scientific claim-evidence extraction and validation, a task that reflects deeper comprehension of scientific argumentation. We systematically compare three approaches which are inspired by divide and conquer approaches, across six diverse LLMs, highlighting model-specific strengths and weaknesses in scientific comprehension. Through evaluation involving over 300 claim-evidence pairs across multiple research domains, we reveal significant limitations in LLMs’ ability to process complex scientific content. Our results demonstrate that closed-source models like GPT-4 and Claude consistently outperform open-source counterparts in precision and recall across claim-evidence identification tasks. Furthermore, strategically designed three-pass and one-by-one prompting approaches significantly improve LLMs’ abilities to accurately link dispersed evidence with claims, although this comes at increased computational cost. CLAIM-BENCH sets a new standard for evaluating scientific comprehension in LLMs, offering both a diagnostic tool and a path forward for building systems capable of deeper, more reliable reasoning across full-length papers.

Large Language Model (LLM)–based research assistant tools demonstrate impressive capabilities, yet their outputs may contain hallucinations that compromise reliability. Therefore, detecting hallucinations in automatically generated scientific content is essential. SciHal2025: Hallucination Detection for Scientific Content challenge @ ACL 2025 provides a valuable platform for advancing this goal. This paper presents our solution to the SciHal2025 challenge. Our method combines several prompting strategies with the fine-tuned base LLMs. We first benchmark multiple LLMs on the SciHal dataset. Next, we developed a detection pipeline that integrates few-shot and chain-of-thought prompting. Hidden representations extracted from the LLMs serve as features for an auxiliary classifier, further improving accuracy. Finally, we fine-tuned the selected base LLMs to enhance end-to-end performance. In this paper, we present comprehensive experimental results and discuss the implications of our findings for future hallucination detection research for scientific content.

Peer assessment is an effective and efficient pedagogical strategy for delivering feedback to learners. Asking students to provide quality feedback, which contains suggestions and mentions problems, can promote metacognition by reviewers and better assist reviewees in revising their work. Thus, various supervised machine learning algorithms have been proposed to detect quality feedback. However, all these powerful algorithms have the same Achilles’ heel: the reliance on sufficient historical data. In other words, collecting adequate peer feedback for training a supervised algorithm can take several semesters before the model can be deployed to a new class. In this paper, we present a new paradigm, called incremental zero-shot learning (IZSL), to tackle the problem of lacking sufficient historical data. Our results show that the method can achieve acceptable “cold-start” performance without needing any domain data, and it outperforms BERT when trained on the same data collected incrementally.