ToolBeHonest: A Multi-level Hallucination Diagnostic Benchmark for Tool-Augmented Large Language Models

Tool-augmented large language models (LLMs) are rapidly being integrated into real-world applications. Due to the lack of benchmarks, the community still needs to fully understand the hallucination issues within these models. To address this challenge, we introduce a comprehensive diagnostic benchmark, ToolBH. Specifically, we assess the LLM's hallucinations through two perspectives: depth and breadth. In terms of depth, we propose a multi-level diagnostic process, including (1) solvability detection, (2) solution planning, and (3) missing-tool analysis. For breadth, we consider three scenarios based on the characteristics of the toolset: missing necessary tools, potential tools, and limited functionality tools. Furthermore, we developed seven tasks and collected 700 evaluation samples through multiple rounds of manual annotation. The results show the significant challenges presented by the ToolBH benchmark. The current advanced models Gemini-1.5-Pro and GPT-4o only achieve a total score of 45.3 and 37.0, respectively, on a scale of 100. In this benchmark, larger model parameters do not guarantee better performance; the training data and response strategies also play a crucial role in tool-enhanced LLM scenarios. Our diagnostic analysis indicates that the primary reason for model errors lies in assessing task solvability. Additionally, open-weight models suffer from performance drops with verbose replies, whereas proprietary models excel with longer reasoning.

PIN: A Knowledge-Intensive Dataset for Paired and Interleaved Multimodal Documents

Recent advancements in Large Multimodal Models (LMMs) have leveraged extensive multimodal datasets to enhance capabilities in complex knowledge-driven tasks. However, persistent challenges in perceptual and reasoning errors limit their efficacy, particularly in interpreting intricate visual data and deducing multimodal relationships. Addressing these issues, we introduce a novel dataset format, PIN (Paired and INterleaved multimodal documents), designed to significantly improve both the depth and breadth of multimodal training. The PIN format is built on three foundational principles: knowledge intensity, scalability, and support for diverse training modalities. This innovative format combines markdown files and comprehensive images to enrich training data with a dense knowledge structure and versatile training strategies. We present PIN-14M, an open-source dataset comprising 14 million samples derived from a diverse range of Chinese and English sources, tailored to include complex web and scientific content. This dataset is constructed meticulously to ensure data quality and ethical integrity, aiming to facilitate advanced training strategies and improve model robustness against common multimodal training pitfalls. Our initial results, forming the basis of this technical report, suggest significant potential for the PIN format in refining LMM performance, with plans for future expansions and detailed evaluations of its impact on model capabilities.

Approximate Angular Domain Expression for Near-Field XL-MIMO Channel

As Extremely Large-Scale Multiple-Input-Multiple-Output (XL-MIMO) technology advances and frequency band rises, the near-field effects in communication are intensifying. A concise and accurate near-field XL-MIMO channel model serves as the cornerstone for investigating the near-field effects. However, existing angular domain XL-MIMO channel models under near-field conditions require non-closed-form wave-number domain integrals for computation, which is complicated. To obtain a more succinct channel model, this paper introduces a closed-form approximate expression based on the principle of stationary phase. It was subsequently shown that when the scatterer distance is larger than the array aperture, the closed-form model can be further simplified as a trapezoidal spectrum. We validate the accuracy of the proposed approximation through simulations of power angular spectrum similarity. The results indicate that the proposed approximation can accurately approximate the near-field angular domain channel within the effective Rayleigh distance.

HoLLMwood: Unleashing the Creativity of Large Language Models in Screenwriting via Role Playing

Generative AI has demonstrated unprecedented creativity in the field of computer vision, yet such phenomena have not been observed in natural language processing. In particular, large language models (LLMs) can hardly produce written works at the level of human experts due to the extremely high complexity of literature writing. In this paper, we present HoLLMwood, an automated framework for unleashing the creativity of LLMs and exploring their potential in screenwriting, which is a highly demanding task. Mimicking the human creative process, we assign LLMs to different roles involved in the real-world scenario. In addition to the common practice of treating LLMs as ${Writer}$, we also apply LLMs as ${Editor}$, who is responsible for providing feedback and revision advice to ${Writer}$. Besides, to enrich the characters and deepen the plots, we introduce a role-playing mechanism and adopt LLMs as ${Actors}$ that can communicate and interact with each other. Evaluations on automatically generated screenplays show that HoLLMwood substantially outperforms strong baselines in terms of coherence, relevance, interestingness and overall quality.

ChartMimic: Evaluating LMM's Cross-Modal Reasoning Capability via Chart-to-Code Generation

We introduce a new benchmark, ChartMimic, aimed at assessing the visually-grounded code generation capabilities of large multimodal models (LMMs). ChartMimic utilizes information-intensive visual charts and textual instructions as inputs, requiring LMMs to generate the corresponding code for chart rendering. ChartMimic includes 1,000 human-curated (figure, instruction, code) triplets, which represent the authentic chart use cases found in scientific papers across various domains(e.g., Physics, Computer Science, Economics, etc). These charts span 18 regular types and 4 advanced types, diversifying into 191 subcategories. Furthermore, we propose multi-level evaluation metrics to provide an automatic and thorough assessment of the output code and the rendered charts. Unlike existing code generation benchmarks, ChartMimic places emphasis on evaluating LMMs' capacity to harmonize a blend of cognitive capabilities, encompassing visual understanding, code generation, and cross-modal reasoning. The evaluation of 3 proprietary models and 11 open-weight models highlights the substantial challenges posed by ChartMimic. Even the advanced GPT-4V, Claude-3-opus only achieve an average score of 73.2 and 53.7, respectively, indicating significant room for improvement. We anticipate that ChartMimic will inspire the development of LMMs, advancing the pursuit of artificial general intelligence.

HHMR: Holistic Hand Mesh Recovery by Enhancing the Multimodal Controllability of Graph Diffusion Models

Recent years have witnessed a trend of the deep integration of the generation and reconstruction paradigms. In this paper, we extend the ability of controllable generative models for a more comprehensive hand mesh recovery task: direct hand mesh generation, inpainting, reconstruction, and fitting in a single framework, which we name as Holistic Hand Mesh Recovery (HHMR). Our key observation is that different kinds of hand mesh recovery tasks can be achieved by a single generative model with strong multimodal controllability, and in such a framework, realizing different tasks only requires giving different signals as conditions. To achieve this goal, we propose an all-in-one diffusion framework based on graph convolution and attention mechanisms for holistic hand mesh recovery. In order to achieve strong control generation capability while ensuring the decoupling of multimodal control signals, we map different modalities to a shared feature space and apply cross-scale random masking in both modality and feature levels. In this way, the correlation between different modalities can be fully exploited during the learning of hand priors. Furthermore, we propose Condition-aligned Gradient Guidance to enhance the alignment of the generated model with the control signals, which significantly improves the accuracy of the hand mesh reconstruction and fitting. Experiments show that our novel framework can realize multiple hand mesh recovery tasks simultaneously and outperform the existing methods in different tasks, which provides more possibilities for subsequent downstream applications including gesture recognition, pose generation, mesh editing, and so on.

Electromagnetic Wave Property Inspired Radio Environment Knowledge Construction and AI-based Verification for 6G Digital Twin Channel

As the underlying foundation of a digital twin network (DTN), a digital twin channel (DTC) can accurately depict the process of radio propagation in the air interface to support the DTN-based 6G wireless network. Since radio propagation is affected by the environment, constructing the relationship between the environment and radio wave propagation is the key to improving the accuracy of DTC, and the construction method based on artificial intelligence (AI) is the most concentrated. However, in the existing methods, the environment information input into the neural network (NN) has many dimensions, and the correlation between the environment and the channel relationship is unclear, resulting in a highly complex relationship construction process. To solve this issue, in this paper, we propose a construction method of radio environment knowledge (REK) inspired by the electromagnetic wave property to quantify the contribution of radio propagation. Specifically, a range selection scheme for effective environment information based on random geometry is proposed to reduce the redundancy of environment information. We quantify the contribution of radio propagation reflection, diffraction and scatterer blockage using environment information and propose a flow chart of REK construction to replace the feature extraction process partially based on NN. To validate REK's effectiveness, we conduct a path loss prediction task based on a lightweight convolutional neural network (CNN) employing a simple two-layer convolutional structure. The results show that the accuracy of the range selection method reaches 90\%; the constructed REK maintains the prediction error of 0.3 and only needs 0.04 seconds of testing time, effectively reducing the network complexity.

4DHands: Reconstructing Interactive Hands in 4D with Transformers

In this paper, we introduce 4DHands, a robust approach to recovering interactive hand meshes and their relative movement from monocular inputs. Our approach addresses two major limitations of previous methods: lacking a unified solution for handling various hand image inputs and neglecting the positional relationship of two hands within images. To overcome these challenges, we develop a transformer-based architecture with novel tokenization and feature fusion strategies. Specifically, we propose a Relation-aware Two-Hand Tokenization (RAT) method to embed positional relation information into the hand tokens. In this way, our network can handle both single-hand and two-hand inputs and explicitly leverage relative hand positions, facilitating the reconstruction of intricate hand interactions in real-world scenarios. As such tokenization indicates the relative relationship of two hands, it also supports more effective feature fusion. To this end, we further develop a Spatio-temporal Interaction Reasoning (SIR) module to fuse hand tokens in 4D with attention and decode them into 3D hand meshes and relative temporal movements. The efficacy of our approach is validated on several benchmark datasets. The results on in-the-wild videos and real-world scenarios demonstrate the superior performances of our approach for interactive hand reconstruction. More video results can be found on the project page: https://4dhands.github.io.

MAP-Neo: Highly Capable and Transparent Bilingual Large Language Model Series

Large Language Models (LLMs) have made great strides in recent years to achieve unprecedented performance across different tasks. However, due to commercial interest, the most competitive models like GPT, Gemini, and Claude have been gated behind proprietary interfaces without disclosing the training details. Recently, many institutions have open-sourced several strong LLMs like LLaMA-3, comparable to existing closed-source LLMs. However, only the model's weights are provided with most details (e.g., intermediate checkpoints, pre-training corpus, and training code, etc.) being undisclosed. To improve the transparency of LLMs, the research community has formed to open-source truly open LLMs (e.g., Pythia, Amber, OLMo), where more details (e.g., pre-training corpus and training code) are being provided. These models have greatly advanced the scientific study of these large models including their strengths, weaknesses, biases and risks. However, we observe that the existing truly open LLMs on reasoning, knowledge, and coding tasks are still inferior to existing state-of-the-art LLMs with similar model sizes. To this end, we open-source MAP-Neo, a highly capable and transparent bilingual language model with 7B parameters trained from scratch on 4.5T high-quality tokens. Our MAP-Neo is the first fully open-sourced bilingual LLM with comparable performance compared to existing state-of-the-art LLMs. Moreover, we open-source all details to reproduce our MAP-Neo, where the cleaned pre-training corpus, data cleaning pipeline, checkpoints, and well-optimized training/evaluation framework are provided. Finally, we hope our MAP-Neo will enhance and strengthen the open research community and inspire more innovations and creativities to facilitate the further improvements of LLMs.

Disttack: Graph Adversarial Attacks Toward Distributed GNN Training

Graph Neural Networks (GNNs) have emerged as potent models for graph learning. Distributing the training process across multiple computing nodes is the most promising solution to address the challenges of ever-growing real-world graphs. However, current adversarial attack methods on GNNs neglect the characteristics and applications of the distributed scenario, leading to suboptimal performance and inefficiency in attacking distributed GNN training. In this study, we introduce Disttack, the first framework of adversarial attacks for distributed GNN training that leverages the characteristics of frequent gradient updates in a distributed system. Specifically, Disttack corrupts distributed GNN training by injecting adversarial attacks into one single computing node. The attacked subgraphs are precisely perturbed to induce an abnormal gradient ascent in backpropagation, disrupting gradient synchronization between computing nodes and thus leading to a significant performance decline of the trained GNN. We evaluate Disttack on four large real-world graphs by attacking five widely adopted GNNs. Compared with the state-of-the-art attack method, experimental results demonstrate that Disttack amplifies the model accuracy degradation by 2.75$\times$ and achieves speedup by 17.33$\times$ on average while maintaining unnoticeability.