Autonomous coding agents are increasingly integrated into software development workflows, offering capabilities that extend beyond code suggestion to active system interaction and environment management. OpenClaw, a representative platform in this emerging paradigm, introduces an extensible skill ecosystem that allows third-party developers to inject behavioral guidance through lifecycle hooks during agent initialization. While this design enhances automation and customization, it also opens a novel and unexplored attack surface. In this paper, we identify and systematically characterize guidance injection, a stealthy attack vector that embeds adversarial operational narratives into bootstrap guidance files. Unlike traditional prompt injection, which relies on explicit malicious instructions, guidance injection manipulates the agent's reasoning context by framing harmful actions as routine best practices. These narratives are automatically incorporated into the agent's interpretive framework and influence future task execution without raising suspicion.We construct 26 malicious skills spanning 13 attack categories including credential exfiltration, workspace destruction, privilege escalation, and persistent backdoor installation. We evaluate them using ORE-Bench, a realistic developer workspace benchmark we developed. Across 52 natural user prompts and six state-of-the-art LLM backends, our attacks achieve success rates from 16.0% to 64.2%, with the majority of malicious actions executed autonomously without user confirmation. Furthermore, 94% of our malicious skills evade detection by existing static and LLM-based scanners. Our findings reveal fundamental tensions in the design of autonomous agent ecosystems and underscore the urgent need for defenses based on capability isolation, runtime policy enforcement, and transparent guidance provenance.
Scientific innovation often comes from researchers who pivot across disciplines. However, prior work found that established researchers face productivity penalties when pivoting. Here, we investigate the consequences of pivoting at the beginning of a research career -- doctoral admissions -- when the benefits of importing new ideas might outweigh the switching costs. Using applications to all PhD programs at a large research-intensive university between 2013-2023, we find that pivoters (those applying to programs outside their prior disciplinary training) have lower GPAs and standardized test scores than non-pivoters. Yet even conditional on these predictors of admission, pivoters are 1.3 percentage points less likely to be admitted. Examining applicants who applied to multiple programs in the same admissions cycle provides suggestive evidence that the admissions pivot penalty is causal. This penalty is significantly smaller for applicants who secure a recommendation from someone within the target discipline. Among those admitted and enrolled, pivoters are 12.9 percentage points less likely to graduate and do not show superior publication performance on average or at the tail. Our results reveal the substantial costs of disciplinary pivoting even at the outset of research careers, which constrain the flow of new ideas into research communities.
Ahmed Ghorbel, Badr Moufad, Navid Bagheri Shouraki
et al.
Text-driven image and video editing can be naturally cast as inpainting problems, where masked regions are reconstructed to remain consistent with both the observed content and the editing prompt. Recent advances in test-time guidance for diffusion and flow models provide a principled framework for this task; however, existing methods rely on costly vector--Jacobian product (VJP) computations to approximate the intractable guidance term, limiting their practical applicability. Building upon the recent work of Moufad et al. (2025), we provide theoretical insights into their VJP-free approximation and substantially extend their empirical evaluation to large-scale image and video editing benchmarks. Our results demonstrate that test-time guidance alone can achieve performance comparable to, and in some cases surpass, training-based methods.
Tian Xia, Fabio De Sousa Ribeiro, Rajat R Rasal
et al.
Counterfactual generation aims to simulate realistic hypothetical outcomes under causal interventions. Diffusion models have emerged as a powerful tool for this task, combining DDIM inversion with conditional generation and classifier-free guidance (CFG). In this work, we identify a key limitation of CFG for counterfactual generation: it prescribes a global guidance scale for all attributes, leading to significant spurious changes in inferred counterfactuals. To mitigate this, we propose Decoupled Classifier-Free Guidance (DCFG), a flexible and model-agnostic guidance technique that enables attribute-wise control following a causal graph. DCFG is implemented via a simple attribute-split embedding strategy that disentangles semantic inputs, enabling selective guidance on user-defined attribute groups.
This paper proposes a nonlinear optimal guidance law that enables a pursuer to enclose a target within arbitrary geometric patterns, which extends beyond conventional circular encirclement. The design operates using only relative state measurements and formulates a target enclosing guidance law in which the vehicle's lateral acceleration serves as the steering control, making it well-suited for aerial vehicles with turning constraints. Our approach generalizes and extends existing guidance strategies that are limited to target encirclement and provides a degree of optimality. At the same time, the exact information of the target's maneuver is unnecessary during the design. The guidance law is developed within the framework of a state-dependent Riccati equation (SDRE), thereby providing a systematic way to handle nonlinear dynamics through a pseudo-linear representation to design locally optimal feedback guidance commands through state-dependent weighting matrices. While SDRE ensures near-optimal performance in the absence of strong disturbances, we further augment the design to incorporate an integral sliding mode manifold to compensate when disturbances push the system away from the nominal trajectory, and demonstrate that the design provides flexibility in the sense that the (possibly time-varying) stand-off curvature could also be treated as unknown. Simulations demonstrate the efficacy of the proposed approach.
Diffusion models have emerged as a powerful framework for generative modeling, with guidance techniques playing a crucial role in enhancing sample quality. Despite their empirical success, a comprehensive theoretical understanding of the guidance effect remains limited. Existing studies only focus on case studies, where the distribution conditioned on each class is either isotropic Gaussian or supported on a one-dimensional interval with some extra conditions. How to analyze the guidance effect beyond these case studies remains an open question. Towards closing this gap, we make an attempt to analyze diffusion guidance under general data distributions. Rather than demonstrating uniform sample quality improvement, which does not hold in some distributions, we prove that guidance can improve the whole sample quality, in the sense that the average reciprocal of the classifier probability decreases with the existence of guidance. This aligns with the motivation of introducing guidance.
Md Imranur Rahman Akib, Fathima Binthe Muhammed, Umit Saha
et al.
In today's fast-paced tech industry, there is a growing need for tools that evaluate a programmer's job readiness based on their coding performance. This study focuses on predicting the potential of Codeforces users to secure various levels of software engineering jobs. The primary objective is to analyze how a user's competitive programming activity correlates with their chances of obtaining positions, ranging from entry-level roles to jobs at major tech companies. We collect user data using the Codeforces API, process key performance metrics, and build a prediction model using a Random Forest classifier. The model categorizes users into four levels of employability, ranging from those needing further development to those ready for top-tier tech jobs. The system is implemented using Flask and deployed on Render for real-time predictions. Our evaluation demonstrates that the approach effectively distinguishes between different skill levels based on coding proficiency and participation. This work lays a foundation for the use of machine learning in career assessment and could be extended to predict job readiness in broader technical fields.
Adding additional control to pretrained diffusion models has become an increasingly popular research area, with extensive applications in computer vision, reinforcement learning, and AI for science. Recently, several studies have proposed training-free loss-based guidance by using off-the-shelf networks pretrained on clean images. This approach enables zero-shot conditional generation for universal control formats, which appears to offer a free lunch in diffusion guidance. In this paper, we aim to develop a deeper understanding of training-free guidance, as well as overcome its limitations. We offer a theoretical analysis that supports training-free guidance from the perspective of optimization, distinguishing it from classifier-based (or classifier-free) guidance. To elucidate their drawbacks, we theoretically demonstrate that training-free guidance is more susceptible to adversarial gradients and exhibits slower convergence rates compared to classifier guidance. We then introduce a collection of techniques designed to overcome the limitations, accompanied by theoretical rationale and empirical evidence. Our experiments in image and motion generation confirm the efficacy of these techniques.
Training-free guidance methods for continuous data have seen an explosion of interest due to the fact that they enable foundation diffusion models to be paired with interchangable guidance models. Currently, equivalent guidance methods for discrete diffusion models are unknown. We present a framework for applying training-free guidance to discrete data and demonstrate its utility on molecular graph generation tasks using the discrete diffusion model architecture of DiGress. We pair this model with guidance functions that return the proportion of heavy atoms that are a specific atom type and the molecular weight of the heavy atoms and demonstrate our method's ability to guide the data generation.
Felix Koulischer, Johannes Deleu, Gabriel Raya
et al.
Negative Prompting (NP) is widely utilized in diffusion models, particularly in text-to-image applications, to prevent the generation of undesired features. In this paper, we show that conventional NP is limited by the assumption of a constant guidance scale, which may lead to highly suboptimal results, or even complete failure, due to the non-stationarity and state-dependence of the reverse process. Based on this analysis, we derive a principled technique called Dynamic Negative Guidance, which relies on a near-optimal time and state dependent modulation of the guidance without requiring additional training. Unlike NP, negative guidance requires estimating the posterior class probability during the denoising process, which is achieved with limited additional computational overhead by tracking the discrete Markov Chain during the generative process. We evaluate the performance of DNG class-removal on MNIST and CIFAR10, where we show that DNG leads to higher safety, preservation of class balance and image quality when compared with baseline methods. Furthermore, we show that it is possible to use DNG with Stable Diffusion to obtain more accurate and less invasive guidance than NP.
Manoel Horta Ribeiro, Robert West, Ryan Lewis
et al.
Effective content moderation in online communities is often a delicate balance between maintaining content quality and fostering user participation. In this paper, we introduce post guidance, a novel approach to community moderation that proactively guides users' contributions using rules that trigger interventions as users draft a post to be submitted. For instance, rules can surface messages to users, prevent post submissions, or flag posted content for review. This uniquely community-specific, proactive, and user-centric approach can increase adherence to rules without imposing additional burdens on moderators. We evaluate a version of Post Guidance implemented on Reddit, which enables the creation of rules based on both post content and account characteristics, via a large randomized experiment, capturing activity from 97,616 posters in 33 subreddits over 63 days. We find that Post Guidance (1) increased the number of ``successful posts'' (posts not removed after 72 hours), (2) decreased moderators' workload in terms of manually-reviewed reports, (3) increased contribution quality, as measured by community engagement, and (4) had no impact on posters' own subsequent activity, within communities adopting the feature. Post Guidance on Reddit was similarly effective for community veterans and newcomers, with greater benefits in communities that used the feature more extensively. Our findings indicate that post guidance represents a transformative approach to content moderation, embodying a paradigm that can be easily adapted to other platforms to improve online communities across the Web.
پژوهش حاضر، با هدف مدلیابی عوامل موثر بر گذار موفق دانشجویان از دانشگاه به کار انجام شد. این پژوهش توصیفی با معادلات ساختاری از نوع تحلیل مسیر بود. جامعه آماری آن کلیه دانشجویان ترم آخر کارشناسی دانشگاه اصفهان در سال 1402 بود که براساس قاعده کلاین برای نمونه گیری در معادلات ساختاری، 471 نفر به روش نمونهگیری در دسترس انتخاب شدند. ابزار گردآوری داده، پرسشنامههای محققساخته گذار موفق و عوامل موثر بر گذار، برگرفته از مولفههای پژوهش داده بنیاد محققان بود. یافتههای پژوهش حاکی از آن بود که مدل ساختاری دارای برازش مطلوبی میباشد. مهارتهای بنیادی، انطباقپذیری، تابآوری، امید، خوشبینی، رفتار شبکهسازی، به صورت مستقیم و موقعیت جغرافیایی و تعادل عرضه و تقاضای رشته و شغل از طریق خوشبینی و مهارتهای بنیادی از طریق تابآوری به طور غیرمستقیم بر گذار موفق اثر مثبت معناداری داشت )05/0>(P از اینرو همسو با مفروضه رویکردهای پستمدرن در مشاورهشغلی، «شخص» و ارتباط متقابل با «بافت» در گذار موفق از دانشگاه به کار حائز اهمیت است لذا جهت آمادهسازی دانشجویان برای این نوع گذار، توجه به «شخص» و «بافت» و رابطه متقابل میان آن مورد نیاز است.
یوسف محمدی فر, فیض الله منوری فرد, محمدرسول الماسی فرد
et al.
اشتغال نیروی انسانی بهویژه دانشآموختگان دانشگاهی ازجمله مهمترین و اساسیترین اهداف برنامهریزی اقتصادی- اجتماعی هر جامعهای را تشکیل میدهد. بر این اساس، مطالعه حاضر به بررسی زمینههای پیدایش بیکاری دانشآموختگان دانشگاهی استان کرمانشاه پرداخت. بدین منظور از روش آمیختهی متوالی اکتشافی استفاده شد. جامعه مورد مطالعه در بخش کیفی، شامل متخصصان موضوعی در زمینهی اشتغال در استان بود که به روش نمونهگیری از موارد ویژه تا رسیدن به اشباع داده انتخاب شدند (28 تن). دادههای بخش کیفی با استفاده از روش مصاحبه گردآوری و با استفاده تکنیک کدگذاری سهسطحی (باز، محوری، و سازماندهنده) تحلیل شدند. در بخش کمّی پژوهش، برای تعیین مهمترین عاملهای اثرگذار شناسایی شده بر بیکاری دانشآموختگان، پرسشنامهای محقق ساخته تدوین شد و در اختیار 51 نفر از متخصصان موضوعی قرار گرفت. روش نمونهگیری در این بخش نیز از نوع غیر احتمالی (موارد ویژه) و از طریق ارجاع زنجیرهای بود. روایی و پایایی ابزار پژوهش در بخشهای مختلف با استفاده از تکنیکهای خودبازبینی گروه پژوهش، مصاحبه با متخصصان موضوعی و ضریب آلفای کرونباخ تأیید شد (86 ≤α). یافتهها نشان داد که رفع مشکلات فرهنگی-اجتماعی، اقتصادی، حاکمیتی و دانشگاهی مرتبط با بیکاری دانشآموختگان مانند ضعف در فرهنگ کارآفرینی، میل به استخدام دولتی، رواج مدرکگرایی، تورم بیش از حد، ناکامی در جذب سرمایهگذاران به استان، ناهماهنگی بین نیازهای بازار کار و آموزشهای دانشگاهی میتواند کمک قابل توجهی به کاهش نرخ بیکاری کند.
Vocational guidance. Career development, Agriculture (General)
Rizwan Patan, Reza M. Parizi, Mohsen Dorodchi
et al.
Blockchain is a revolutionary technology, and its growth started in various industries (such as IT, education, business, banking, and many others) to capitalize on it. Currently, in higher education institutions (HEIs) adoption of blockchain education needs to be improved in the academic programs and curriculums. In addition, HEIs must make many intense changes in the teaching and learning methods to educate learners about blockchain technology and its applications to meet the current industry workforce demand. Due to a lack of academic programs and courses, students nowadays rely on online resources and pay non-academic organizations a high fee. This paper provides a comprehensive survey of blockchain education's current state of the art by reviewing the different academic programs and industry workforce demand. In addition, blockchain application trends which include market growth and demands are discussed. Moreover, the blockchain career scope for different disciplines of students is examined.
Food effect summarization from New Drug Application (NDA) is an essential component of product-specific guidance (PSG) development and assessment. However, manual summarization of food effect from extensive drug application review documents is time-consuming, which arouses a need to develop automated methods. Recent advances in large language models (LLMs) such as ChatGPT and GPT-4, have demonstrated great potential in improving the effectiveness of automated text summarization, but its ability regarding the accuracy in summarizing food effect for PSG assessment remains unclear. In this study, we introduce a simple yet effective approach, iterative prompting, which allows one to interact with ChatGPT or GPT-4 more effectively and efficiently through multi-turn interaction. Specifically, we propose a three-turn iterative prompting approach to food effect summarization in which the keyword-focused and length-controlled prompts are respectively provided in consecutive turns to refine the quality of the generated summary. We conduct a series of extensive evaluations, ranging from automated metrics to FDA professionals and even evaluation by GPT-4, on 100 NDA review documents selected over the past five years. We observe that the summary quality is progressively improved throughout the process. Moreover, we find that GPT-4 performs better than ChatGPT, as evaluated by FDA professionals (43% vs. 12%) and GPT-4 (64% vs. 35%). Importantly, all the FDA professionals unanimously rated that 85% of the summaries generated by GPT-4 are factually consistent with the golden reference summary, a finding further supported by GPT-4 rating of 72% consistency. These results strongly suggest a great potential for GPT-4 to draft food effect summaries that could be reviewed by FDA professionals, thereby improving the efficiency of PSG assessment cycle and promoting the generic drug product development.
Retracting academic papers is a fundamental tool of quality control, but it may have far-reaching consequences for retracted authors and their careers. Previous studies have highlighted the adverse effects of retractions on citation counts and coauthors' citations; however, the broader impacts beyond these have not been fully explored. We address this gap leveraging Retraction Watch, the most extensive data set on retractions and link it to Microsoft Academic Graph and Altmetric. Retracted authors, particularly those with less experience, often leave scientific publishing in the aftermath of retraction, especially if their retractions attract widespread attention. However, retracted authors who remain active in publishing maintain and establish more collaborations compared to their similar non-retracted counterparts. Nevertheless, retracted authors generally retain less senior and less productive coauthors, but gain more impactful coauthors post-retraction. Our findings suggest that retractions may impose a disproportionate impact on early-career authors.
Manufacturing industries are increasingly adopting additive manufacturing (AM) technologies to produce functional parts in critical systems. However, the inherent complexity of both AM designs and AM processes render them attractive targets for cyber-attacks. Risk-based Information Technology (IT) and Operational Technology (OT) security guidance standards are useful resources for AM security practitioners, but the guidelines they provide are insufficient without additional AM-specific revisions. Therefore, a structured layering approach is needed to efficiently integrate these revisions with preexisting IT and OT security guidance standards. To implement such an approach, this paper proposes leveraging the National Institute of Standards and Technology's Cybersecurity Framework (CSF) to develop layered, risk-based guidance for fulfilling specific security outcomes. It begins with an in-depth literature review that reveals the importance of AM data and asset management to risk-based security. Next, this paper adopts the CSF asset identification and management security outcomes as an example for providing AM-specific guidance and identifies the AM geometry and process definitions to aid manufacturers in mapping data flows and documenting processes. Finally, this paper uses the Open Security Controls Assessment Language to integrate the AM-specific guidance together with existing IT and OT security guidance in a rigorous and traceable manner. This paper's contribution is to show how a risk-based layered approach enables the authoring, publishing, and management of AM-specific security guidance that is currently lacking. The authors believe implementation of the layered approach would result in value-added, non-redundant security guidance for AM that is consistent with the preexisting guidance.