We introduce ClaimCheck, an LLM-guided automatic fact-checking system designed to verify real-world claims using live Web evidence and small language models. Unlike prior systems that rely on large, closed-source models and static knowledge stores, ClaimCheck employs a transparent, stepwise verification pipeline that mirrors human fact-checking workflows consisting of Web search query planning, Web-based evidence retrieval and summarization, evidence synthesis and re-retrieval, and claim verdict evaluation. Each module is optimized for small LLMs, allowing the system to deliver accurate and interpretable fact-checking with significantly lower computational requirements. Despite using a much smaller Qwen3-4B model, ClaimCheck achieves state-of-the-art accuracy of 76.4% on the AVeriTeC dataset, outperforming previous approaches using LLaMA3.1 70B and GPT-4o. Extensive ablations demonstrate that careful modular design and prompting strategies can overcome the limitations of smaller LLMs. To promote accessibility and transparency, we provide a public demo at https://idir.uta.edu/claimcheck.

Diversity is an essential metric for evaluating the creativity of outputs generated by language models. Temperature-based sampling is a common strategy to increase diversity. However, for tasks that require high precision, e.g., mathematical reasoning, uncontrolled high temperature sampling, e.g., min-$p$ or top-$p$, degrades reasoning quality. We demonstrate that the loss of accuracy is caused by sampling incorrect continuations in sensitive decoding positions. To address this, in this paper, we propose \textbf{selective sampling}, a method that dynamically switches between greedy and high-temperature sampling based on a sampling risk metric. This risk metric estimates the likelihood of output errors when applying high-temperature sampling on the current token position. To predict sampling risk, we train a lightweight classifier on a small subset of verifiable problems. The trained classifier can be integrated with the base language model with minimal latency overhead. Experiments on mathematical reasoning tasks demonstrate that selective sampling enhances the quality-diversity trade-off, even in high-temperature settings.

The stock market is extremely difficult to predict in the short term due to high market volatility, changes caused by news, and the non-linear nature of the financial time series. This research proposes a novel framework for improving minute-level prediction accuracy using semantic sentiment scores from top ten different large language models (LLMs) combined with minute interval intraday stock price data. We systematically constructed a time-aligned dataset of AAPL news articles and 1-minute Apple Inc. (AAPL) stock prices for the dates of April 4 to May 2, 2025. The sentiment analysis was achieved using the DeepSeek-V3, GPT variants, LLaMA, Claude, Gemini, Qwen, and Mistral models through their APIs. Each article obtained sentiment scores from all ten LLMs, which were scaled to a [0, 1] range and combined with prices and technical indicators like RSI, ROC, and Bollinger Band Width. Two state-of-the-art such as Reformer and Mamba were trained separately on the dataset using the sentiment scores produced by each LLM as input. Hyper parameters were optimized by means of Optuna and were evaluated through a 3-day evaluation period. Reformer had mean squared error (MSE) or the evaluation metrics, and it should be noted that Mamba performed not only faster but also better than Reformer for every LLM across the 10 LLMs tested. Mamba performed best with LLaMA 3.3--70B, with the lowest error of 0.137. While Reformer could capture broader trends within the data, the model appeared to over smooth sudden changes by the LLMs. This study highlights the potential of integrating LLM-based semantic analysis paired with efficient temporal modeling to enhance real-time financial forecasting.

GPT can predict users' future decisions by analyzing narrative tables, with accuracy further improved when guided by an anthropological framework. Moreover, by integrating contextual knowledge and an interpretative lens into LLMs, this approach enhances AI explainability while ensuring a human-centric perspective in requirement elicitation. By asking GPT to generate a user profile, it becomes possible to directly assess what the model has understood about the user and how it represents them. Furthermore, since the model is not only tasked with predicting users' responses in new scenarios but also with justifying its choices, it is possible, on one hand, to understand the rationale behind the model's output and, on the other, to identify potential misalignments between the model's prediction and the user's actual values and preferences. This enables targeted interventions to improve alignment between the LLM and the user profile, creating a continuous feedback loop that involves both the user and the LLM trained to interpret data through an anthropological lens. The process strengthens the model's interpretability, ethical alignment, and predictive adaptability, thereby making AI systems more transparent and attuned to real-world human values. Ultimately, the approach lays the groundwork for AI assistants capable of recognizing and adapting to individuals' soft ethics and ethical decision-making process. B. Threat to V alidity We discuss threats to validity following the qualitative research framework proposed in [72]--namely, credibility, transferability, dependability, and confirmability.

Large-scale models are pretrained on massive web-crawled datasets containing documents of mixed quality, making data filtering essential. A popular method is Classifier-based Quality Filtering (CQF), which trains a binary classifier to distinguish between pretraining data and a small, high-quality set. It assigns each pretraining document a quality score defined as the classifier's score and retains only the top-scoring ones. We provide an in-depth analysis of CQF. We show that while CQF improves downstream task performance, it does not necessarily enhance language modeling on the high-quality dataset. We explain this paradox by the fact that CQF implicitly filters the high-quality dataset as well. We further compare the behavior of models trained with CQF to those trained on synthetic data of increasing quality, obtained via random token permutations, and find starkly different trends. Our results challenge the view that CQF captures a meaningful notion of data quality.

The development of trustworthy artificial intelligence requires moving beyond black-box performance metrics toward an understanding of models' internal computations. Mechanistic Interpretability (MI) aims to meet this need by identifying the algorithmic mechanisms underlying model behaviors. Yet, the scientific rigor of MI critically depends on the reliability of its findings. In this work, we argue that interpretability methods, such as circuit discovery, should be viewed as statistical estimators, subject to questions of variance and robustness. To illustrate this statistical framing, we present a systematic stability analysis of a state-of-the-art circuit discovery method: EAP-IG. We evaluate its variance and robustness through a comprehensive suite of controlled perturbations, including input resampling, prompt paraphrasing, hyperparameter variation, and injected noise within the causal analysis itself. Across a diverse set of models and tasks, our results demonstrate that EAP-IG exhibits high structural variance and sensitivity to hyperparameters, questioning the stability of its findings. Based on these results, we offer a set of best-practice recommendations for the field, advocating for the routine reporting of stability metrics to promote a more rigorous and statistically grounded science of interpretability.

Time Series Foundation Models (TSFMs) have shown promising zero-shot generalization across diverse forecasting tasks. However, their robustness to continual adaptation remains underexplored. In this work, we investigate the extent to which TSFMs suffer from catastrophic forgetting when fine-tuned sequentially on multiple datasets. Using synthetic datasets designed with varying degrees of periodic structure, we measure the trade-off between adaptation to new data and retention of prior knowledge. Our experiments reveal that, while fine-tuning improves performance on new tasks, it often causes significant degradation on previously learned ones, illustrating a fundamental stability-plasticity dilemma.

Inherently, robotic manipulation tasks are history-dependent: leveraging past context could be beneficial. However, most existing Vision-Language-Action models (VLAs) have been designed without considering this aspect, i.e., they rely solely on the current observation, ignoring preceding context. In this paper, we propose HAMLET, a scalable framework to adapt VLAs to attend to the historical context during action prediction. Specifically, we introduce moment tokens that compactly encode perceptual information at each timestep. Their representations are initialized with time-contrastive learning, allowing them to better capture temporally distinctive aspects. Next, we employ a lightweight memory module that integrates the moment tokens across past timesteps into memory features, which are then leveraged for action prediction. Through empirical evaluation, we show that HAMLET successfully transforms a state-of-the-art VLA into a history-aware policy, especially demonstrating significant improvements on long-horizon tasks that require historical context. In particular, on top of GR00T N1.5, HAMLET achieves an average success rate of 76.4% on history-dependent real-world tasks, surpassing the baseline performance by 47.2%. Furthermore, HAMLET pushes prior art performance from 64.1% to 66.4% on RoboCasa Kitchen (100-demo setup) and from 95.6% to 97.7% on LIBERO, highlighting its effectiveness even under generic robot-manipulation benchmarks.

Recent advances in reasoning capabilities of large language models (LLMs) are largely driven by reinforcement learning (RL), yet the underlying parameter dynamics during RL training remain poorly understood. This work identifies two fundamental properties of RL-induced parameter updates in LLMs: (1) Rank-1 Dominance, where the top singular subspace of the parameter update matrix nearly fully determines reasoning improvements, recovering over 99\% of performance gains; and (2) Rank-1 Linear Dynamics, where this dominant subspace evolves linearly throughout training, enabling accurate prediction from early checkpoints. Extensive experiments across 8 LLMs and 7 algorithms validate the generalizability of these properties. More importantly, based on these findings, we propose AlphaRL, a plug-in acceleration framework that extrapolates the final parameter update using a short early training window, achieving up to 2.5 speedup while retaining \textgreater 96\% of reasoning performance without extra modules or hyperparameter tuning. This positions our finding as a versatile and practical tool for large-scale RL, opening a path toward principled, interpretable, and efficient training paradigm for LLMs.

Machine learning models for clinical prediction rely on structured data extracted from Electronic Medical Records (EMRs), yet this process remains dominated by hardcoded, database-specific pipelines for cohort definition, feature selection, and code mapping. These manual efforts limit scalability, reproducibility, and cross-institutional generalization. To address this, we introduce EMR-AGENT (Automated Generalized Extraction and Navigation Tool), an agent-based framework that replaces manual rule writing with dynamic, language model-driven interaction to extract and standardize structured clinical data. Our framework automates cohort selection, feature extraction, and code mapping through interactive querying of databases. Our modular agents iteratively observe query results and reason over schema and documentation, using SQL not just for data retrieval but also as a tool for database observation and decision making. This eliminates the need for hand-crafted, schema-specific logic. To enable rigorous evaluation, we develop a benchmarking codebase for three EMR databases (MIMIC-III, eICU, SICdb), including both seen and unseen schema settings. Our results demonstrate strong performance and generalization across these databases, highlighting the feasibility of automating a process previously thought to require expert-driven design. The code will be released publicly at https://github.com/AITRICS/EMR-AGENT/tree/main. For a demonstration, please visit our anonymous demo page: https://anonymoususer-max600.github.io/EMR_AGENT/