With the recent advancement of Large Language Models (LLMs), generating functionally correct code has become less complicated for a wide array of developers. While using LLMs has sped up the functional development process, it poses a heavy risk to code security. Code generation with proper security measures using LLM is a significantly more challenging task than functional code generation. Security measures may include adding a pair of lines of code with the original code, consisting of null pointer checking or prepared statements for SQL injection prevention. Currently, available code repair LLMs generate code repair by supervised fine-tuning, where the model looks at cross-entropy loss. However, the original and repaired codes are mostly similar in functionality and syntactically, except for a few (1-2) lines, which act as security measures. This imbalance between the lines needed for security measures and the functional code enforces the supervised fine-tuned model to prioritize generating functional code without adding proper security measures, which also benefits the model by resulting in minimal loss. Therefore, in this work, for security hardening and strengthening of generated code from LLMs, we propose a reinforcement learning-based method for program-specific repair with the combination of semantic and syntactic reward mechanisms that focus heavily on adding security and functional measures in the code, respectively.
相關內容
With the rapid development of Large Language Models (LLMs), various explorations have arisen to utilize LLMs capability of context understanding on recommender systems. While pioneering strategies have primarily transformed traditional recommendation tasks into challenges of natural language generation, there has been a relative scarcity of exploration in the domain of session-based recommendation (SBR) due to its specificity. SBR has been primarily dominated by Graph Neural Networks, which have achieved many successful outcomes due to their ability to capture both the implicit and explicit relationships between adjacent behaviors. The structural nature of graphs contrasts with the essence of natural language, posing a significant adaptation gap for LLMs. In this paper, we introduce large language models with graphical Session-Based recommendation, named LLMGR, an effective framework that bridges the aforementioned gap by harmoniously integrating LLMs with Graph Neural Networks (GNNs) for SBR tasks. This integration seeks to leverage the complementary strengths of LLMs in natural language understanding and GNNs in relational data processing, leading to a more powerful session-based recommender system that can understand and recommend items within a session. Moreover, to endow the LLM with the capability to empower SBR tasks, we design a series of prompts for both auxiliary and major instruction tuning tasks. These prompts are crafted to assist the LLM in understanding graph-structured data and align textual information with nodes, effectively translating nuanced user interactions into a format that can be understood and utilized by LLM architectures. Extensive experiments on three real-world datasets demonstrate that LLMGR outperforms several competitive baselines, indicating its effectiveness in enhancing SBR tasks and its potential as a research direction for future exploration.
Kernel ridge regression, KRR, is a generalization of linear ridge regression that is non-linear in the data, but linear in the parameters. Here, we introduce an equivalent formulation of the objective function of KRR, opening up both for using penalties other than the ridge penalty and for studying kernel ridge regression from the perspective of gradient descent. Using a continuous-time perspective, we derive a closed-form solution for solving kernel regression with gradient descent, something we refer to as kernel gradient flow, KGF, and theoretically bound the differences between KRR and KGF, where, for the latter, regularization is obtained through early stopping. We also generalize KRR by replacing the ridge penalty with the $\ell_1$ and $\ell_\infty$ penalties, respectively, and use the fact that analogous to the similarities between KGF and KRR, $\ell_1$ regularization and forward stagewise regression (also known as coordinate descent), and $\ell_\infty$ regularization and sign gradient descent, follow similar solution paths. We can thus alleviate the need for computationally heavy algorithms based on proximal gradient descent. We show theoretically and empirically how the $\ell_1$ and $\ell_\infty$ penalties, and the corresponding gradient-based optimization algorithms, produce sparse and robust kernel regression solutions, respectively.
Bilingual Lexicon Induction (BLI) is a core task in multilingual NLP that still, to a large extent, relies on calculating cross-lingual word representations. Inspired by the global paradigm shift in NLP towards Large Language Models (LLMs), we examine the potential of the latest generation of LLMs for the development of bilingual lexicons. We ask the following research question: Is it possible to prompt and fine-tune multilingual LLMs (mLLMs) for BLI, and how does this approach compare against and complement current BLI approaches? To this end, we systematically study 1) zero-shot prompting for unsupervised BLI and 2) few-shot in-context prompting with a set of seed translation pairs, both without any LLM fine-tuning, as well as 3) standard BLI-oriented fine-tuning of smaller LLMs. We experiment with 18 open-source text-to-text mLLMs of different sizes (from 0.3B to 13B parameters) on two standard BLI benchmarks covering a range of typologically diverse languages. Our work is the first to demonstrate strong BLI capabilities of text-to-text mLLMs. The results reveal that few-shot prompting with in-context examples from nearest neighbours achieves the best performance, establishing new state-of-the-art BLI scores for many language pairs. We also conduct a series of in-depth analyses and ablation studies, providing more insights on BLI with (m)LLMs, also along with their limitations.
Traditional Reinforcement Learning (RL) algorithms are usually applied in robotics to learn controllers that act with a fixed control rate. Given the discrete nature of RL algorithms, they are oblivious to the effects of the choice of control rate: finding the correct control rate can be difficult and mistakes often result in excessive use of computing resources or even lack of convergence. We propose Soft Elastic Actor-Critic (SEAC), a novel off-policy actor-critic algorithm to address this issue. SEAC implements elastic time steps, time steps with a known, variable duration, which allow the agent to change its control frequency to adapt to the situation. In practice, SEAC applies control only when necessary, minimizing computational resources and data usage. We evaluate SEAC's capabilities in simulation in a Newtonian kinematics maze navigation task and on a 3D racing video game, Trackmania. SEAC outperforms the SAC baseline in terms of energy efficiency and overall time management, and most importantly without the need to identify a control frequency for the learned controller. SEAC demonstrated faster and more stable training speeds than SAC, especially at control rates where SAC struggled to converge. We also compared SEAC with a similar approach, the Continuous-Time Continuous-Options (CTCO) model, and SEAC resulted in better task performance. These findings highlight the potential of SEAC for practical, real-world RL applications in robotics.
The incredible success of transformers on sequence modeling tasks can be largely attributed to the self-attention mechanism, which allows information to be transferred between different parts of a sequence. Self-attention allows transformers to encode causal structure which makes them particularly suitable for sequence modeling. However, the process by which transformers learn such causal structure via gradient-based training algorithms remains poorly understood. To better understand this process, we introduce an in-context learning task that requires learning latent causal structure. We prove that gradient descent on a simplified two-layer transformer learns to solve this task by encoding the latent causal graph in the first attention layer. The key insight of our proof is that the gradient of the attention matrix encodes the mutual information between tokens. As a consequence of the data processing inequality, the largest entries of this gradient correspond to edges in the latent causal graph. As a special case, when the sequences are generated from in-context Markov chains, we prove that transformers learn an induction head (Olsson et al., 2022). We confirm our theoretical findings by showing that transformers trained on our in-context learning task are able to recover a wide variety of causal structures.
Considering a conversation thread, rumour stance classification aims to identify the opinion (e.g. agree or disagree) of replies towards a target (rumour story). Although the target is expected to be an essential component in traditional stance classification, we show that rumour stance classification datasets contain a considerable amount of real-world data whose stance could be naturally inferred directly from the replies, contributing to the strong performance of the supervised models without awareness of the target. We find that current target-aware models underperform in cases where the context of the target is crucial. Finally, we propose a simple yet effective framework to enhance reasoning with the targets, achieving state-of-the-art performance on two benchmark datasets.
We propose a noble, comprehensive and robust agile requirements change management (ARCM) model that addresses the limitations of existing models and is tailored for agile software development in the global software development paradigm. To achieve this goal, we conducted an exhaustive literature review and an empirical study with RCM industry experts. Our study evaluated the effectiveness of the proposed RCM model in a real-world setting and identifies any limitations or areas for improvement. The results of our study provide valuable insights into how the proposed RCM model can be applied in agile global software development environments to improve software development practices and optimize project success rates.
Experimental data in Particle and Nuclear physics, Particle Astrophysics and Radiation Protection Dosimetry are obtained from experimental facilities comprising a complex array of sensors, electronics and software. Computer simulation is used to study the measurement process. Probability Density Functions (PDFs) of measured physical parameters deviate from true PDFs due to resolution, bias, and efficiency effects. Good estimates of the true PDF are necessary for testing theoretical models, comparing results from different experiments, and combining results from various research endeavors. In the article, the histogram method is employed to estimate both the measured and true PDFs. The binning of histograms is determined using the K-means clustering algorithm. The true PDF is estimated through the maximization of the likelihood function with entropy regularization, utilizing a non-linear optimization algorithm specially designed for this purpose. The accuracy of the results is assessed using the Mean Integrated Square Error. To determine the optimal value for the regularization parameter, a bootstrap method is applied. Additionally, a mathematical model of the measurement system is formulated using system identification methods. This approach enhances the robustness and precision of the estimation process, providing a more reliable analysis of the system's characteristics.
Graph Neural Networks (GNNs) have shown promising results on a broad spectrum of applications. Most empirical studies of GNNs directly take the observed graph as input, assuming the observed structure perfectly depicts the accurate and complete relations between nodes. However, graphs in the real world are inevitably noisy or incomplete, which could even exacerbate the quality of graph representations. In this work, we propose a novel Variational Information Bottleneck guided Graph Structure Learning framework, namely VIB-GSL, in the perspective of information theory. VIB-GSL advances the Information Bottleneck (IB) principle for graph structure learning, providing a more elegant and universal framework for mining underlying task-relevant relations. VIB-GSL learns an informative and compressive graph structure to distill the actionable information for specific downstream tasks. VIB-GSL deduces a variational approximation for irregular graph data to form a tractable IB objective function, which facilitates training stability. Extensive experimental results demonstrate that the superior effectiveness and robustness of VIB-GSL.
In LiDAR-based 3D object detection for autonomous driving, the ratio of the object size to input scene size is significantly smaller compared to 2D detection cases. Overlooking this difference, many 3D detectors directly follow the common practice of 2D detectors, which downsample the feature maps even after quantizing the point clouds. In this paper, we start by rethinking how such multi-stride stereotype affects the LiDAR-based 3D object detectors. Our experiments point out that the downsampling operations bring few advantages, and lead to inevitable information loss. To remedy this issue, we propose Single-stride Sparse Transformer (SST) to maintain the original resolution from the beginning to the end of the network. Armed with transformers, our method addresses the problem of insufficient receptive field in single-stride architectures. It also cooperates well with the sparsity of point clouds and naturally avoids expensive computation. Eventually, our SST achieves state-of-the-art results on the large scale Waymo Open Dataset. It is worth mentioning that our method can achieve exciting performance (83.8 LEVEL 1 AP on validation split) on small object (pedestrian) detection due to the characteristic of single stride. Codes will be released at //github.com/TuSimple/SST
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191