Autoformalization aims to convert informal mathematical proofs into machine-verifiable formats, bridging the gap between natural and formal languages. However, ensuring semantic alignment between the informal and formalized statements remains challenging. Existing approaches heavily rely on manual verification, hindering scalability. To address this, we introduce \textsc{FormalAlign}, the first automated framework designed for evaluating the alignment between natural and formal languages in autoformalization. \textsc{FormalAlign} trains on both the autoformalization sequence generation task and the representational alignment between input and output, employing a dual loss that combines a pair of mutually enhancing autoformalization and alignment tasks. Evaluated across four benchmarks augmented by our proposed misalignment strategies, \textsc{FormalAlign} demonstrates superior performance. In our experiments, \textsc{FormalAlign} outperforms GPT-4, achieving an Alignment-Selection Score 11.58\% higher on \forml-Basic (99.21\% vs. 88.91\%) and 3.19\% higher on MiniF2F-Valid (66.39\% vs. 64.34\%). This effective alignment evaluation significantly reduces the need for manual verification. Both the dataset and code can be accessed via~\url{https://github.com/rookie-joe/FormalAlign}.

In human cognition theory, human thinking is governed by two systems: the fast and intuitive System 1 and the slower but more deliberative System 2. Recent studies have shown that incorporating System 2 process into Transformers including large language models (LLMs), significantly enhances their reasoning capabilities. Nevertheless, models that purely resemble System 2 thinking require substantially higher computational costs and are much slower to respond. To address this challenge, we present Dualformer, a single Transformer model that seamlessly integrates both the fast and slow reasoning modes. Dualformer is obtained by training on data with randomized reasoning traces, where different parts of the traces are dropped during training. The dropping strategies are specifically tailored according to the trace structure, analogous to analyzing our thinking process and creating shortcuts with patterns. At inference time, our model can be configured to output only the solutions (fast mode) or both the reasoning chain and the final solution (slow mode), or automatically decide which mode to engage (auto mode). In all cases, Dualformer outperforms the corresponding baseline models in both performance and computational efficiency: (1) in slow mode, Dualformer optimally solves unseen 30 x 30 maze navigation tasks 97.6% of the time, surpassing the Searchformer (trained on data with complete reasoning traces) baseline performance of 93.3%, while only using 45.5% fewer reasoning steps; (2) in fast mode, Dualformer completes those tasks with an 80% optimal rate, significantly outperforming the Solution-Only model (trained on solution-only data), which has an optimal rate of only 30%. For math problems, our techniques have also achieved improved performance with LLM fine-tuning, showing its generalization beyond task-specific models.

While much research has focused on hallucinations in multiple modalities including images and natural language text, less attention has been given to hallucinations in source code, which leads to incorrect and vulnerable code that causes significant financial loss. To pave the way for research in LLMs' hallucinations in code, we introduce Collu-Bench, a benchmark for predicting code hallucinations of LLMs across code generation (CG) and automated program repair (APR) tasks. Collu-Bench includes 13,234 code hallucination instances collected from five datasets and 11 diverse LLMs, ranging from open-source models to commercial ones. To better understand and predict code hallucinations, Collu-Bench provides detailed features such as the per-step log probabilities of LLMs' output, token types, and the execution feedback of LLMs' generated code for in-depth analysis. In addition, we conduct experiments to predict hallucination on Collu-Bench, using both traditional machine learning techniques and neural networks, which achieves 22.03 - 33.15% accuracy. Our experiments draw insightful findings of code hallucination patterns, reveal the challenge of accurately localizing LLMs' hallucinations, and highlight the need for more sophisticated techniques. Despite the great potential and impressive success of LLMs (Touvron et al., 2023; Brown et al., 2020; Li et al., 2022a; OpenAI, 2024), a known issue of LLMs is hallucination, a phenomenon where the model generates fluent and plausible-sounding but unfaithful or fabricated content (Ji et al., 2023). The hallucination issue poses a significant risk when deploying LLMs in real-world applications that require precise information (Puchert et al., 2023). Due to this importance, researchers have developed benchmarks such as TruthfulQA (Lin et al., 2022), FELM (chen et al., 2023), and HaluEval (Li et al., 2023b) to understand and predict hallucinations of LLMs. Additionally, researchers are actively exploring methods to mitigate hallucinations (Liu et al., 2024b; Elaraby et al., 2023; Dhuliawala et al., 2023; Yan et al., 2024). Another domain where LLMs have been widely applied is source code.

This paper investigates the design of a unified search engine to serve multiple retrieval-augmented generation (RAG) agents, each with a distinct task, backbone large language model (LLM), and retrieval-augmentation strategy. We introduce an iterative approach where the search engine generates retrieval results for these RAG agents and gathers feedback on the quality of the retrieved documents during an offline phase. This feedback is then used to iteratively optimize the search engine using a novel expectation-maximization algorithm, with the goal of maximizing each agent's utility function. Additionally, we adapt this approach to an online setting, allowing the search engine to refine its behavior based on real-time individual agents feedback to better serve the results for each of them. Experiments on diverse datasets from the Knowledge-Intensive Language Tasks (KILT) benchmark demonstrates that our approach significantly on average outperforms competitive baselines across 18 RAG models. We also demonstrate that our method effectively ``personalizes'' the retrieval process for each RAG agent based on the collected feedback. Finally, we provide a comprehensive ablation study to explore various aspects of our method.

This paper presents an innovative large language model (LLM) agent framework for enhancing diagnostic accuracy in simulated clinical environments using the AgentClinic benchmark. The proposed automatic correction enables doctor agents to iteratively refine their reasoning and actions following incorrect diagnoses, fostering improved decision-making over time. Experiments show that the implementation of the adaptive LLM-based doctor agents achieve correct diagnoses through dynamic interactions with simulated patients. The evaluations highlight the capacity of autonomous agents to adapt and improve in complex medical scenarios. Future enhancements will focus on refining the algorithm and expanding its applicability across a wider range of tasks and different large language models.

Conversation agents powered by large language models are revolutionizing the way we interact with visual data. Recently, large vision-language models (LVLMs) have been extensively studied for both images and videos. However, these studies typically focus on common scenarios. In this work, we introduce an LVLM specifically designed for surgical scenarios. We integrate visual representations of surgical images and videos into the language feature space. Consequently, we establish a LVLM model, Surgical-LLaVA, fine-tuned on instruction following data of surgical scenarios. Our experiments demonstrate that Surgical-LLaVA exhibits impressive multi-modal chat abilities in surgical contexts, occasionally displaying multi-modal behaviors on unseen instructions. We conduct a quantitative evaluation of visual question-answering datasets for surgical scenarios. The results show superior performance compared to previous works, indicating the potential of our model to tackle more complex surgery scenarios.

Fine-tuning is powerful for adapting large language models to downstream tasks, but it often results in huge memory usages. A promising approach to mitigate this is using Zeroth-Order (ZO) optimization, which estimates gradients to replace First-Order (FO) gradient calculations, albeit with longer training time due to its stochastic nature. By revisiting the Memory-efficient ZO (MeZO) optimizer, we discover that the full-parameter perturbation and updating processes consume over 50% of its overall fine-tuning time cost. Based on these observations, we introduce a novel layer-wise sparse computation and memory efficient ZO optimizer, named LeZO. LeZO treats layers as fundamental units for sparsification and dynamically perturbs different parameter subsets in each step to achieve full-parameter fine-tuning. LeZO incorporates layer-wise parameter sparsity in the process of simultaneous perturbation stochastic approximation (SPSA) and ZO stochastic gradient descent (ZO-SGD). It achieves accelerated computation during perturbation and updating processes without additional memory overhead. We conduct extensive experiments with the OPT model family on the SuperGLUE benchmark and two generative tasks. The experiments show that LeZO accelerates training without compromising the performance of ZO optimization. Specifically, it achieves over 3x speedup compared to MeZO on the SST-2, BoolQ, and Copa tasks.

Large Language Models (LLMs) have exhibited remarkable capabilities in many complex tasks including mathematical reasoning. However, traditional approaches heavily rely on ensuring self-consistency within single prompting method, which limits the exploration of diverse problem-solving strategies. This study addresses these limitations by performing an experimental analysis of distinct prompting methods within the domain of mathematical reasoning. Our findings demonstrate that each method explores a distinct search space, and this differentiation becomes more evident with increasing problem complexity. To leverage this phenomenon, we applied efficient sampling process that uniformly combines samples from these diverse methods, which not only expands the maximum search space but achieves higher performance with fewer runs compared to single methods. Especially, within the subset of difficult questions of MATH dataset named MATH-hard, The maximum search space was achieved while utilizing approximately 43% fewer runs than single methods on average. These findings highlight the importance of integrating diverse problem-solving strategies to enhance the reasoning abilities of LLMs.

Parameter-efficient fine-tuning methods, such as Low-Rank Adaptation (LoRA), are known to enhance training efficiency in Large Language Models (LLMs). Due to the limited parameters of LoRA, recent studies seek to combine LoRA with Mixture-of-Experts (MoE) to boost performance across various tasks. However, inspired by the observed redundancy in traditional MoE structures, previous studies identify similar redundancy among LoRA experts within the MoE architecture, highlighting the necessity for non-uniform allocation of LoRA experts across different layers. In this paper, we leverage Heavy-Tailed Self-Regularization (HT-SR) Theory to design a fine-grained allocation strategy. Our analysis reveals that the number of experts per layer correlates with layer training quality, which exhibits significant variability across layers. Based on this, we introduce AlphaLoRA, a theoretically principled and training-free method for allocating LoRA experts to further mitigate redundancy. Experiments on three models across ten language processing and reasoning benchmarks demonstrate that AlphaLoRA achieves comparable or superior performance over all baselines. Our code is available at https://github.com/morelife2017/alphalora.

Cyber-physical systems like autonomous vehicles are tested in simulation before deployment, using domain-specific programs for scenario specification. To aid the testing of autonomous vehicles in simulation, we design a natural language interface, using an instruction-following large language model, to assist a non-coding domain expert in synthesising the desired scenarios and vehicle behaviours. We show that using it to convert utterances to the symbolic program is feasible, despite the very small training dataset. Human experiments show that dialogue is critical to successful simulation generation, leading to a 4.5 times higher success rate than a generation without engaging in extended conversation.