Numerous applications of large language models (LLMs) rely on their ability to perform step-by-step reasoning. However, the reasoning behavior of LLMs remains poorly understood, posing challenges to research, development, and safety. To address this gap, we introduce landscape of thoughts-the first visualization tool for users to inspect the reasoning paths of chain-of-thought and its derivatives on any multi-choice dataset. Specifically, we represent the states in a reasoning path as feature vectors that quantify their distances to all answer choices. These features are then visualized in two-dimensional plots using t-SNE. Qualitative and quantitative analysis with the landscape of thoughts effectively distinguishes between strong and weak models, correct and incorrect answers, as well as different reasoning tasks. It also uncovers undesirable reasoning patterns, such as low consistency and high uncertainty. Additionally, users can adapt our tool to a model that predicts the property they observe. We showcase this advantage by adapting our tool to a lightweight verifier that evaluates the correctness of reasoning paths. The code is publicly available at: https://github.com/tmlr-group/landscape-of-thoughts.

Lane keeping in autonomous driving systems requires scenario-specific weight tuning for different objectives. We formulate lane-keeping as a constrained reinforcement learning problem, where weight coefficients are automatically learned along with the policy, eliminating the need for scenario-specific tuning. Empirically, our approach outperforms traditional RL in efficiency and reliability. Additionally, real-world demonstrations validate its practical value for real-world autonomous driving.

Graph neural networks, recently introduced into the field of fluid flow surrogate modeling, have been successfully applied to model the temporal evolution of various fluid flow systems. Existing applications, however, are mostly restricted to cases where the domain is time-invariant. The present work extends the application of graph neural network-based modeling to fluid flow around structures rotating with respect to a certain axis. Specifically, we propose to apply a graph neural network-based surrogate modeling for fluid flow with the mesh corotating with the structure. Unlike conventional data-driven approaches that rely on structured Cartesian meshes, our framework operates on unstructured co-rotating meshes, enforcing rotation equivariance of the learned model by leveraging co-rotating polar (2D) and cylindrical (3D) coordinate systems. To model the pressure for systems without Dirichlet pressure boundaries, we propose a novel local directed pressure difference formulation that is invariant to the reference pressure point and value. For flow systems with large mesh sizes, we introduce a scheme to train the network in single or distributed graphics processing units by accumulating the backpropagated gradients from partitions of the mesh. The effectiveness of our proposed framework is examined on two test cases: (i) fluid flow in a 2D rotating mixer, and (ii) the flow past a 3D rotating cube. Our results show that the model achieves stable and accurate rollouts for over 2000 time steps in periodic regimes while capturing accurate short-term dynamics in chaotic flow regimes. In addition, the drag and lift force predictions closely match the CFD calculations, highlighting the potential of the framework for modeling both periodic and chaotic fluid flow around rotating structures.

This survey examines evaluation methods for large language model (LLM)-based agents in multi-turn conversational settings. Using a PRISMA-inspired framework, we systematically reviewed nearly 250 scholarly sources, capturing the state of the art from various venues of publication, and establishing a solid foundation for our analysis. Our study offers a structured approach by developing two interrelated taxonomy systems: one that defines \emph{what to evaluate} and another that explains \emph{how to evaluate}. The first taxonomy identifies key components of LLM-based agents for multi-turn conversations and their evaluation dimensions, including task completion, response quality, user experience, memory and context retention, as well as planning and tool integration. These components ensure that the performance of conversational agents is assessed in a holistic and meaningful manner. The second taxonomy system focuses on the evaluation methodologies. It categorizes approaches into annotation-based evaluations, automated metrics, hybrid strategies that combine human assessments with quantitative measures, and self-judging methods utilizing LLMs. This framework not only captures traditional metrics derived from language understanding, such as BLEU and ROUGE scores, but also incorporates advanced techniques that reflect the dynamic, interactive nature of multi-turn dialogues.

The increasing emphasis on privacy and data security has driven the adoption of federated learning, a decentralized approach to train machine learning models without sharing raw data. Prompt learning, which fine-tunes prompt embeddings of pretrained models, offers significant advantages in federated settings by reducing computational costs and communication overheads while leveraging the strong performance and generalization capabilities of vision-language models such as CLIP. This paper addresses the intersection of federated learning and prompt learning, particularly for vision-language models. In this work, we introduce a comprehensive framework, named FLIP, to evaluate federated prompt learning algorithms. FLIP assesses the performance of 8 state-of-the-art federated prompt learning methods across 4 federated learning protocols and 12 open datasets, considering 6 distinct evaluation scenarios. Our findings demonstrate that prompt learning maintains strong generalization performance in both in-distribution and out-of-distribution settings with minimal resource consumption. This work highlights the effectiveness of federated prompt learning in environments characterized by data scarcity, unseen classes, and cross-domain distributional shifts. We open-source the code for all implemented algorithms in FLIP to facilitate further research in this domain.

Common factors and microcounseling skills are critical to the effectiveness of psychotherapy. Understanding and measuring these elements provides valuable insights into therapeutic processes and outcomes. However, automatic identification of these change principles from textual data remains challenging due to the nuanced and context-dependent nature of therapeutic dialogue. This paper introduces CFiCS, a hierarchical classification framework integrating graph machine learning with pretrained contextual embeddings. We represent common factors, intervention concepts, and microcounseling skills as a heterogeneous graph, where textual information from ClinicalBERT enriches each node. This structure captures both the hierarchical relationships (e.g., skill-level nodes linking to broad factors) and the semantic properties of therapeutic concepts. By leveraging graph neural networks, CFiCS learns inductive node embeddings that generalize to unseen text samples lacking explicit connections. Our results demonstrate that integrating ClinicalBERT node features and graph structure significantly improves classification performance, especially in fine-grained skill prediction. CFiCS achieves substantial gains in both micro and macro F1 scores across all tasks compared to baselines, including random forests, BERT-based multi-task models, and graph-based methods.

The extraction of geoelectric structural information from airborne transient electromagnetic(ATEM)data primarily involves data processing and inversion. Conventional methods rely on empirical parameter selection, making it difficult to process complex field data with high noise levels. Additionally, inversion computations are time consuming and often suffer from multiple local minima. Existing deep learning-based approaches separate the data processing steps, where independently trained denoising networks struggle to ensure the reliability of subsequent inversions. Moreover, end to end networks lack interpretability. To address these issues, we propose a unified and interpretable deep learning inversion paradigm based on disentangled representation learning. The network explicitly decomposes noisy data into noise and signal factors, completing the entire data processing workflow based on the signal factors while incorporating physical information for guidance. This approach enhances the network's reliability and interpretability. The inversion results on field data demonstrate that our method can directly use noisy data to accurately reconstruct the subsurface electrical structure. Furthermore, it effectively processes data severely affected by environmental noise, which traditional methods struggle with, yielding improved lateral structural resolution.

The 3D Gaussian Splatting (3D-GS) is a novel method for scene representation and view synthesis. Although Scaffold-GS achieves higher quality real-time rendering compared to the original 3D-GS, its fine-grained rendering of the scene is extremely dependent on adequate viewing angles. The spectral bias of neural network learning results in Scaffold-GS's poor ability to perceive and learn high-frequency information in the scene. In this work, we propose enhancing the manifold complexity of input features and using network-based feature map loss to improve the image reconstruction quality of 3D-GS models. We introduce AH-GS, which enables 3D Gaussians in structurally complex regions to obtain higher-frequency encodings, allowing the model to more effectively learn the high-frequency information of the scene. Additionally, we incorporate high-frequency reinforce loss to further enhance the model's ability to capture detailed frequency information. Our result demonstrates that our model significantly improves rendering fidelity, and in specific scenarios (e.g., MipNeRf360-garden), our method exceeds the rendering quality of Scaffold-GS in just 15K iterations.

Motivated in part by their relevance for low-precision training and quantization, massive activations in large language models (LLMs) have recently emerged as a topic of interest. However, existing analyses are limited in scope, and generalizability across architectures is unclear. This paper helps address some of these gaps by conducting an analysis of massive activations across a broad range of LLMs, including both GLU-based and non-GLU-based architectures. Our findings challenge several prior assumptions, most importantly: (1) not all massive activations are detrimental, i.e. suppressing them does not lead to an explosion of perplexity or a collapse in downstream task performance; (2) proposed mitigation strategies such as Attention KV bias are model-specific and ineffective in certain cases. We consequently investigate novel hybrid mitigation strategies; in particular pairing Target Variance Rescaling (TVR) with Attention KV bias or Dynamic Tanh (DyT) successfully balances the mitigation of massive activations with preserved downstream model performance in the scenarios we investigated. Our code is available at: https://github.com/bluorion-com/refine_massive_activations.

The rapid advancement of large language models (LLMs) has introduced new challenges in distinguishing human-written text from AI-generated content. In this work, we explored a pipelined approach for AI-generated text detection that includes a feature extraction step (i.e. prompt-based rewriting features inspired by RAIDAR and content-based features derived from the NELA toolkit) followed by a classification module. Comprehensive experiments were conducted on the Defactify4.0 dataset, evaluating two tasks: binary classification to differentiate human-written and AI-generated text, and multi-class classification to identify the specific generative model used to generate the input text. Our findings reveal that NELA features significantly outperform RAIDAR features in both tasks, demonstrating their ability to capture nuanced linguistic, stylistic, and content-based differences. Combining RAIDAR and NELA features provided minimal improvement, highlighting the redundancy introduced by less discriminative features. Among the classifiers tested, XGBoost emerged as the most effective, leveraging the rich feature sets to achieve high accuracy and generalisation.