On Path to Multimodal Historical Reasoning: HistBench and HistAgent

Recent advances in large language models (LLMs) have led to remarkable progress across domains, yet their capabilities in the humanities, particularly history, remain underexplored. Historical reasoning poses unique challenges for AI, involving multimodal source interpretation, temporal inference, and cross-linguistic analysis. While general-purpose agents perform well on many existing benchmarks, they lack the domain-specific expertise required to engage with historical materials and questions. To address this gap, we introduce HistBench, a new benchmark of 414 high-quality questions designed to evaluate AI's capacity for historical reasoning and authored by more than 40 expert contributors. The tasks span a wide range of historical problems-from factual retrieval based on primary sources to interpretive analysis of manuscripts and images, to interdisciplinary challenges involving archaeology, linguistics, or cultural history. Furthermore, the benchmark dataset spans 29 ancient and modern languages and covers a wide range of historical periods and world regions. Finding the poor performance of LLMs and other agents on HistBench, we further present HistAgent, a history-specific agent equipped with carefully designed tools for OCR, translation, archival search, and image understanding in History. On HistBench, HistAgent based on GPT-4o achieves an accuracy of 27.54% pass@1 and 36.47% pass@2, significantly outperforming LLMs with online search and generalist agents, including GPT-4o (18.60%), DeepSeek-R1(14.49%) and Open Deep Research-smolagents(20.29% pass@1 and 25.12% pass@2). These results highlight the limitations of existing LLMs and generalist agents and demonstrate the advantages of HistAgent for historical reasoning.

Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution

Recent advances in large language models (LLMs) have enabled agents to autonomously perform complex, open-ended tasks. However, many existing frameworks depend heavily on manually predefined tools and workflows, which hinder their adaptability, scalability, and generalization across domains. In this work, we introduce Alita--a generalist agent designed with the principle of "Simplicity is the ultimate sophistication," enabling scalable agentic reasoning through minimal predefinition and maximal self-evolution. For minimal predefinition, Alita is equipped with only one component for direct problem-solving, making it much simpler and neater than previous approaches that relied heavily on hand-crafted, elaborate tools and workflows. This clean design enhances its potential to generalize to challenging questions, without being limited by tools. For Maximal self-evolution, we enable the creativity of Alita by providing a suite of general-purpose components to autonomously construct, refine, and reuse external capabilities by generating task-related model context protocols (MCPs) from open source, which contributes to scalable agentic reasoning. Notably, Alita achieves 75.15% pass@1 and 87.27% pass@3 accuracy, which is top-ranking among general-purpose agents, on the GAIA benchmark validation dataset, 74.00% and 52.00% pass@1, respectively, on Mathvista and PathVQA, outperforming many agent systems with far greater complexity. More details will be updated at $\href{https://github.com/CharlesQ9/Alita}{https://github.com/CharlesQ9/Alita}$.

Shallow Preference Signals: Large Language Model Aligns Even Better with Truncated Data?

Aligning large language models (LLMs) with human preferences remains a key challenge in AI. Preference-based optimization methods, such as Reinforcement Learning with Human Feedback (RLHF) and Direct Preference Optimization (DPO), rely on human-annotated datasets to improve alignment. In this work, we identify a crucial property of the existing learning method: the distinguishing signal obtained in preferred responses is often concentrated in the early tokens. We refer to this as shallow preference signals. To explore this property, we systematically truncate preference datasets at various points and train both reward models and DPO models on the truncated data. Surprisingly, models trained on truncated datasets, retaining only the first half or fewer tokens, achieve comparable or even superior performance to those trained on full datasets. For example, a reward model trained on the Skywork-Reward-Preference-80K-v0.2 dataset outperforms the full dataset when trained on a 40\% truncated dataset. This pattern is consistent across multiple datasets, suggesting the widespread presence of shallow preference signals. We further investigate the distribution of the reward signal through decoding strategies. We consider two simple decoding strategies motivated by the shallow reward signal observation, namely Length Control Decoding and KL Threshold Control Decoding, which leverage shallow preference signals to optimize the trade-off between alignment and computational efficiency. The performance is even better, which again validates our hypothesis. The phenomenon of shallow preference signals highlights potential issues in LLM alignment: existing alignment methods often focus on aligning only the initial tokens of responses, rather than considering the full response. This could lead to discrepancies with real-world human preferences, resulting in suboptimal alignment performance.

A Systematic Survey of Automatic Prompt Optimization Techniques

Since the advent of large language models (LLMs), prompt engineering has been a crucial step for eliciting desired responses for various Natural Language Processing (NLP) tasks. However, prompt engineering remains an impediment for end users due to rapid advances in models, tasks, and associated best practices. To mitigate this, Automatic Prompt Optimization (APO) techniques have recently emerged that use various automated techniques to help improve the performance of LLMs on various tasks. In this paper, we present a comprehensive survey summarizing the current progress and remaining challenges in this field. We provide a formal definition of APO, a 5-part unifying framework, and then proceed to rigorously categorize all relevant works based on their salient features therein. We hope to spur further research guided by our framework.

DebateQA: Evaluating Question Answering on Debatable Knowledge

The rise of large language models (LLMs) has enabled us to seek answers to inherently debatable questions on LLM chatbots, necessitating a reliable way to evaluate their ability. However, traditional QA benchmarks assume fixed answers are inadequate for this purpose. To address this, we introduce DebateQA, a dataset of 2,941 debatable questions, each accompanied by multiple human-annotated partial answers that capture a variety of perspectives. We develop two metrics: Perspective Diversity, which evaluates the comprehensiveness of perspectives, and Dispute Awareness, which assesses if the LLM acknowledges the question's debatable nature. Experiments demonstrate that both metrics align with human preferences and are stable across different underlying models. Using DebateQA with two metrics, we assess 12 popular LLMs and retrieval-augmented generation methods. Our findings reveal that while LLMs generally excel at recognizing debatable issues, their ability to provide comprehensive answers encompassing diverse perspectives varies considerably.

Truthful Dataset Valuation by Pointwise Mutual Information

A common way to evaluate a dataset in ML involves training a model on this dataset and assessing the model's performance on a test set. However, this approach has two issues: (1) it may incentivize undesirable data manipulation in data marketplaces, as the self-interested data providers seek to modify the dataset to maximize their evaluation scores; (2) it may select datasets that overfit to potentially small test sets. We propose a new data valuation method that provably guarantees the following: data providers always maximize their expected score by truthfully reporting their observed data. Any manipulation of the data, including but not limited to data duplication, adding random data, data removal, or re-weighting data from different groups, cannot increase their expected score. Our method, following the paradigm of proper scoring rules, measures the pointwise mutual information (PMI) of the test dataset and the evaluated dataset. However, computing the PMI of two datasets is challenging. We introduce a novel PMI measuring method that greatly improves tractability within Bayesian machine learning contexts. This is accomplished through a new characterization of PMI that relies solely on the posterior probabilities of the model parameter at an arbitrarily selected value. Finally, we support our theoretical results with simulations and further test the effectiveness of our data valuation method in identifying the top datasets among multiple data providers. Interestingly, our method outperforms the standard approach of selecting datasets based on the trained model's test performance, suggesting that our truthful valuation score can also be more robust to overfitting.

The Evolution of the Interplay Between Input Distributions and Linear Regions in Networks

It is commonly recognized that the expressiveness of deep neural networks is contingent upon a range of factors, encompassing their depth, width, and other relevant considerations. Currently, the practical performance of the majority of deep neural networks remains uncertain. For ReLU (Rectified Linear Unit) networks with piecewise linear activations, the number of linear convex regions serves as a natural metric to gauge the network's expressivity. In this paper, we count the number of linear convex regions in deep neural networks based on ReLU. In particular, we prove that for any one-dimensional input, there exists a minimum threshold for the number of neurons required to express it. We also empirically observe that for the same network, intricate inputs hinder its capacity to express linear regions. Furthermore, we unveil the iterative refinement process of decision boundaries in ReLU networks during training. We aspire for our research to serve as an inspiration for network optimization endeavors and aids in the exploration and analysis of the behaviors exhibited by deep networks.