Effective Training Data Synthesis for Improving MLLM Chart Understanding

Being able to effectively read scientific plots, or chart understanding, is a central part toward building effective agents for science. However, existing multimodal large language models (MLLMs), especially open-source ones, are still falling behind with a typical success rate of 30%-50% on challenging benchmarks. Previous studies on fine-tuning MLLMs with synthetic charts are often restricted by their inadequate similarity to the real charts, which could compromise model training and performance on complex real-world charts. In this study, we show that modularizing chart generation and diversifying visual details improves chart understanding capabilities. In particular, we design a five-step data synthesis pipeline, where we separate data and function creation for single plot generation, condition the generation of later subplots on earlier ones for multi-subplot figures, visually diversify the generated figures, filter out low quality data, and finally generate the question-answer (QA) pairs with GPT-4o. This approach allows us to streamline the generation of fine-tuning datasets and introduce the effective chart dataset (ECD), which contains 10k+ chart images and 300k+ QA pairs, covering 25 topics and featuring 250+ chart type combinations with high visual complexity. We show that ECD consistently improves the performance of various MLLMs on a range of real-world and synthetic test sets. Code, data and models are available at: https://github.com/yuweiyang-anu/ECD.

A Reproducible, Scalable Pipeline for Synthesizing Autoregressive Model Literature

The accelerating pace of research on autoregressive generative models has produced thousands of papers, making manual literature surveys and reproduction studies increasingly impractical. We present a fully open-source, reproducible pipeline that automatically retrieves candidate documents from public repositories, filters them for relevance, extracts metadata, hyper-parameters and reported results, clusters topics, produces retrieval-augmented summaries and generates containerised scripts for re-running selected experiments. Quantitative evaluation on 50 manually-annotated papers shows F1 scores above 0.85 for relevance classification, hyper-parameter extraction and citation identification. Experiments on corpora of up to 1000 papers demonstrate near-linear scalability with eight CPU workers. Three case studies -- AWD-LSTM on WikiText-2, Transformer-XL on WikiText-103 and an autoregressive music model on the Lakh MIDI dataset -- confirm that the extracted settings support faithful reproduction, achieving test perplexities within 1--3% of the original reports.

ChatGPT on the Road: Leveraging Large Language Model-Powered In-vehicle Conversational Agents for Safer and More Enjoyable Driving Experience

Studies on in-vehicle conversational agents have traditionally relied on pre-scripted prompts or limited voice commands, constraining natural driver-agent interaction. To resolve this issue, the present study explored the potential of a ChatGPT-based in-vehicle agent capable of carrying continuous, multi-turn dialogues. Forty drivers participated in our experiment using a motion-based driving simulator, comparing three conditions (No agent, Pre-scripted agent, and ChatGPT-based agent) as a within-subjects variable. Results showed that the ChatGPT-based agent condition led to more stable driving performance across multiple metrics. Participants demonstrated lower variability in longitudinal acceleration, lateral acceleration, and lane deviation compared to the other two conditions. In subjective evaluations, the ChatGPT-based agent also received significantly higher ratings in competence, animacy, affective trust, and preference compared to the Pre-scripted agent. Our thematic analysis of driver-agent conversations revealed diverse interaction patterns in topics, including driving assistance/questions, entertainment requests, and anthropomorphic interactions. Our results highlight the potential of LLM-powered in-vehicle conversational agents to enhance driving safety and user experience through natural, context-rich interactions.

The Problem with Safety Classification is not just the Models

Studying the robustness of Large Language Models (LLMs) to unsafe behaviors is an important topic of research today. Building safety classification models or guard models, which are fine-tuned models for input/output safety classification for LLMs, is seen as one of the solutions to address the issue. Although there is a lot of research on the safety testing of LLMs themselves, there is little research on evaluating the effectiveness of such safety classifiers or the evaluation datasets used for testing them, especially in multilingual scenarios. In this position paper, we demonstrate how multilingual disparities exist in 5 safety classification models by considering datasets covering 18 languages. At the same time, we identify potential issues with the evaluation datasets, arguing that the shortcomings of current safety classifiers are not only because of the models themselves. We expect that these findings will contribute to the discussion on developing better methods to identify harmful content in LLM inputs across languages.

Interpretable Topic Extraction and Word Embedding Learning using row-stochastic DEDICOM

The DEDICOM algorithm provides a uniquely interpretable matrix factorization method for symmetric and asymmetric square matrices. We employ a new row-stochastic variation of DEDICOM on the pointwise mutual information matrices of text corpora to identify latent topic clusters within the vocabulary and simultaneously learn interpretable word embeddings. We introduce a method to efficiently train a constrained DEDICOM algorithm and a qualitative evaluation of its topic modeling and word embedding performance.

S2WTM: Spherical Sliced-Wasserstein Autoencoder for Topic Modeling

Modeling latent representations in a hyperspherical space has proven effective for capturing directional similarities in high-dimensional text data, benefiting topic modeling. Variational autoencoder-based neural topic models (VAE-NTMs) commonly adopt the von Mises-Fisher prior to encode hyperspherical structure. However, VAE-NTMs often suffer from posterior collapse, where the KL divergence term in the objective function highly diminishes, leading to ineffective latent representations. To mitigate this issue while modeling hyperspherical structure in the latent space, we propose the Spherical Sliced Wasserstein Autoencoder for Topic Modeling (S2WTM). S2WTM employs a prior distribution supported on the unit hypersphere and leverages the Spherical Sliced-Wasserstein distance to align the aggregated posterior distribution with the prior. Experimental results demonstrate that S2WTM outperforms state-of-the-art topic models, generating more coherent and diverse topics while improving performance on downstream tasks.

CLAP: Coreference-Linked Augmentation for Passage Retrieval

Large Language Model (LLM)-based passage expansion has shown promise for enhancing first-stage retrieval, but often underperforms with dense retrievers due to semantic drift and misalignment with their pretrained semantic space. Beyond this, only a portion of a passage is typically relevant to a query, while the rest introduces noise--an issue compounded by chunking techniques that break coreference continuity. We propose Coreference-Linked Augmentation for Passage Retrieval (CLAP), a lightweight LLM-based expansion framework that segments passages into coherent chunks, resolves coreference chains, and generates localized pseudo-queries aligned with dense retriever representations. A simple fusion of global topical signals and fine-grained subtopic signals achieves robust performance across domains. CLAP yields consistent gains even as retriever strength increases, enabling dense retrievers to match or surpass second-stage rankers such as BM25 + MonoT5-3B, with up to 20.68% absolute nDCG@10 improvement. These improvements are especially notable in out-of-domain settings, where conventional LLM-based expansion methods relying on domain knowledge often falter. CLAP instead adopts a logic-centric pipeline that enables robust, domain-agnostic generalization.

NGTM: Substructure-based Neural Graph Topic Model for Interpretable Graph Generation

Graph generation plays a pivotal role across numerous domains, including molecular design and knowledge graph construction. Although existing methods achieve considerable success in generating realistic graphs, their interpretability remains limited, often obscuring the rationale behind structural decisions. To address this challenge, we propose the Neural Graph Topic Model (NGTM), a novel generative framework inspired by topic modeling in natural language processing. NGTM represents graphs as mixtures of latent topics, each defining a distribution over semantically meaningful substructures, which facilitates explicit interpretability at both local and global scales. The generation process transparently integrates these topic distributions with a global structural variable, enabling clear semantic tracing of each generated graph. Experiments demonstrate that NGTM achieves competitive generation quality while uniquely enabling fine-grained control and interpretability, allowing users to tune structural features or induce biological properties through topic-level adjustments.

Characterizing Deep Research: A Benchmark and Formal Definition

Information tasks such as writing surveys or analytical reports require complex search and reasoning, and have recently been grouped under the umbrella of \textit{deep research} -- a term also adopted by recent models targeting these capabilities. Despite growing interest, the scope of the deep research task remains underdefined and its distinction from other reasoning-intensive problems is poorly understood. In this paper, we propose a formal characterization of the deep research (DR) task and introduce a benchmark to evaluate the performance of DR systems. We argue that the core defining feature of deep research is not the production of lengthy report-style outputs, but rather the high fan-out over concepts required during the search process, i.e., broad and reasoning-intensive exploration. To enable objective evaluation, we define DR using an intermediate output representation that encodes key claims uncovered during search-separating the reasoning challenge from surface-level report generation. Based on this formulation, we propose a diverse, challenging benchmark LiveDRBench with 100 challenging tasks over scientific topics (e.g., datasets, materials discovery, prior art search) and public interest events (e.g., flight incidents, movie awards). Across state-of-the-art DR systems, F1 score ranges between 0.02 and 0.72 for any sub-category. OpenAI's model performs the best with an overall F1 score of 0.55. Analysis of reasoning traces reveals the distribution over the number of referenced sources, branching, and backtracking events executed by current DR systems, motivating future directions for improving their search mechanisms and grounding capabilities. The benchmark is available at https://github.com/microsoft/LiveDRBench.

Watch the Weights: Unsupervised monitoring and control of fine-tuned LLMs

The releases of powerful open-weight large language models (LLMs) are often not accompanied by access to their full training data. Existing interpretability methods, particularly those based on activations, often require or assume distributionally similar data. This is a significant limitation when detecting and defending against novel potential threats like backdoors, which are by definition out-of-distribution. In this work, we introduce a new method for understanding, monitoring and controlling fine-tuned LLMs that interprets weights, rather than activations, thereby side stepping the need for data that is distributionally similar to the unknown training data. We demonstrate that the top singular vectors of the weight difference between a fine-tuned model and its base model correspond to newly acquired behaviors. By monitoring the cosine similarity of activations along these directions, we can detect salient behaviors introduced during fine-tuning with high precision. For backdoored models that bypasses safety mechanisms when a secret trigger is present, our method stops up to 100% of attacks with a false positive rate below 1.2%. For models that have undergone unlearning, we detect inference on erased topics with accuracy up to 95.42% and can even steer the model to recover "unlearned" information. Besides monitoring, our method also shows potential for pre-deployment model auditing: by analyzing commercial instruction-tuned models (OLMo, Llama, Qwen), we are able to uncover model-specific fine-tuning focus including marketing strategies and Midjourney prompt generation. Our implementation can be found at https://github.com/fjzzq2002/WeightWatch.