Introducing the Generative Application Firewall (GAF)cs.AI updates on arXiv.org arXiv:2601.15824v2 Announce Type: replace-cross
Abstract: This paper introduces the Generative Application Firewall (GAF), a new architectural layer for securing LLM applications. Existing defenses — prompt filters, guardrails, and data-masking — remain fragmented; GAF unifies them into a single enforcement point, much like a WAF coordinates defenses for web traffic, while also covering autonomous agents and their tool interactions.
arXiv:2601.15824v2 Announce Type: replace-cross
Abstract: This paper introduces the Generative Application Firewall (GAF), a new architectural layer for securing LLM applications. Existing defenses — prompt filters, guardrails, and data-masking — remain fragmented; GAF unifies them into a single enforcement point, much like a WAF coordinates defenses for web traffic, while also covering autonomous agents and their tool interactions. Read More
Exploring LLMs for Scientific Information Extraction Using The SciEx Frameworkcs.AI updates on arXiv.org arXiv:2512.10004v2 Announce Type: replace
Abstract: Large language models (LLMs) are increasingly touted as powerful tools for automating scientific information extraction. However, existing methods and tools often struggle with the realities of scientific literature: long-context documents, multi-modal content, and reconciling varied and inconsistent fine-grained information across multiple publications into standardized formats. These challenges are further compounded when the desired data schema or extraction ontology changes rapidly, making it difficult to re-architect or fine-tune existing systems. We present SciEx, a modular and composable framework that decouples key components including PDF parsing, multi-modal retrieval, extraction, and aggregation. This design streamlines on-demand data extraction while enabling extensibility and flexible integration of new models, prompting strategies, and reasoning mechanisms. We evaluate SciEx on datasets spanning three scientific topics for its ability to extract fine-grained information accurately and consistently. Our findings provide practical insights into both the strengths and limitations of current LLM-based pipelines.
arXiv:2512.10004v2 Announce Type: replace
Abstract: Large language models (LLMs) are increasingly touted as powerful tools for automating scientific information extraction. However, existing methods and tools often struggle with the realities of scientific literature: long-context documents, multi-modal content, and reconciling varied and inconsistent fine-grained information across multiple publications into standardized formats. These challenges are further compounded when the desired data schema or extraction ontology changes rapidly, making it difficult to re-architect or fine-tune existing systems. We present SciEx, a modular and composable framework that decouples key components including PDF parsing, multi-modal retrieval, extraction, and aggregation. This design streamlines on-demand data extraction while enabling extensibility and flexible integration of new models, prompting strategies, and reasoning mechanisms. We evaluate SciEx on datasets spanning three scientific topics for its ability to extract fine-grained information accurately and consistently. Our findings provide practical insights into both the strengths and limitations of current LLM-based pipelines. Read More
LLM is Not All You Need: A Systematic Evaluation of ML vs. Foundation Models for text and image based Medical Classificationcs.AI updates on arXiv.org arXiv:2601.16549v1 Announce Type: new
Abstract: The combination of multimodal Vision-Language Models (VLMs) and Large Language Models (LLMs) opens up new possibilities for medical classification. This work offers a rigorous, unified benchmark by using four publicly available datasets covering text and image modalities (binary and multiclass complexity) that contrasts traditional Machine Learning (ML) with contemporary transformer-based techniques. We evaluated three model classes for each task: Classical ML (LR, LightGBM, ResNet-50), Prompt-Based LLMs/VLMs (Gemini 2.5), and Fine-Tuned PEFT Models (LoRA-adapted Gemma3 variants). All experiments used consistent data splits and aligned metrics. According to our results, traditional machine learning (ML) models set a high standard by consistently achieving the best overall performance across most medical categorization tasks. This was especially true for structured text-based datasets, where the classical models performed exceptionally well. In stark contrast, the LoRA-tuned Gemma variants consistently showed the worst performance across all text and image experiments, failing to generalize from the minimal fine-tuning provided. However, the zero-shot LLM/VLM pipelines (Gemini 2.5) had mixed results; they performed poorly on text-based tasks, but demonstrated competitive performance on the multiclass image task, matching the classical ResNet-50 baseline. These results demonstrate that in many medical categorization scenarios, established machine learning models continue to be the most reliable option. The experiment suggests that foundation models are not universally superior and that the effectiveness of Parameter-Efficient Fine-Tuning (PEFT) is highly dependent on the adaptation strategy, as minimal fine-tuning proved detrimental in this study.
arXiv:2601.16549v1 Announce Type: new
Abstract: The combination of multimodal Vision-Language Models (VLMs) and Large Language Models (LLMs) opens up new possibilities for medical classification. This work offers a rigorous, unified benchmark by using four publicly available datasets covering text and image modalities (binary and multiclass complexity) that contrasts traditional Machine Learning (ML) with contemporary transformer-based techniques. We evaluated three model classes for each task: Classical ML (LR, LightGBM, ResNet-50), Prompt-Based LLMs/VLMs (Gemini 2.5), and Fine-Tuned PEFT Models (LoRA-adapted Gemma3 variants). All experiments used consistent data splits and aligned metrics. According to our results, traditional machine learning (ML) models set a high standard by consistently achieving the best overall performance across most medical categorization tasks. This was especially true for structured text-based datasets, where the classical models performed exceptionally well. In stark contrast, the LoRA-tuned Gemma variants consistently showed the worst performance across all text and image experiments, failing to generalize from the minimal fine-tuning provided. However, the zero-shot LLM/VLM pipelines (Gemini 2.5) had mixed results; they performed poorly on text-based tasks, but demonstrated competitive performance on the multiclass image task, matching the classical ResNet-50 baseline. These results demonstrate that in many medical categorization scenarios, established machine learning models continue to be the most reliable option. The experiment suggests that foundation models are not universally superior and that the effectiveness of Parameter-Efficient Fine-Tuning (PEFT) is highly dependent on the adaptation strategy, as minimal fine-tuning proved detrimental in this study. Read More
AR-LIF: Adaptive reset leaky integrate-and-fire neuron for spiking neural networkscs.AI updates on arXiv.org arXiv:2507.20746v3 Announce Type: replace-cross
Abstract: Spiking neural networks offer low energy consumption due to their event-driven nature. Beyond binary spike outputs, their intrinsic floating-point dynamics merit greater attention. Neuronal threshold levels and reset modes critically determine spike count and timing. Hard reset cause information loss, while soft reset apply uniform treatment to neurons. To address these issues, we design an adaptive reset neuron that establishes relationships between inputs, outputs, and reset, while integrating a simple yet effective threshold adjustment strategy. Experimental results demonstrate that our method achieves excellent performance while maintaining lower energy consumption. In particular, it attains state-of-the-art accuracy on Tiny-ImageNet and CIFAR10-DVS. Codes are available at https://github.com/2ephyrus/AR-LIF.
arXiv:2507.20746v3 Announce Type: replace-cross
Abstract: Spiking neural networks offer low energy consumption due to their event-driven nature. Beyond binary spike outputs, their intrinsic floating-point dynamics merit greater attention. Neuronal threshold levels and reset modes critically determine spike count and timing. Hard reset cause information loss, while soft reset apply uniform treatment to neurons. To address these issues, we design an adaptive reset neuron that establishes relationships between inputs, outputs, and reset, while integrating a simple yet effective threshold adjustment strategy. Experimental results demonstrate that our method achieves excellent performance while maintaining lower energy consumption. In particular, it attains state-of-the-art accuracy on Tiny-ImageNet and CIFAR10-DVS. Codes are available at https://github.com/2ephyrus/AR-LIF. Read More
LUMINA: Long-horizon Understanding for Multi-turn Interactive Agentscs.AI updates on arXiv.org arXiv:2601.16649v1 Announce Type: new
Abstract: Large language models can perform well on many isolated tasks, yet they continue to struggle on multi-turn, long-horizon agentic problems that require skills such as planning, state tracking, and long context processing. In this work, we aim to better understand the relative importance of advancing these underlying capabilities for success on such tasks. We develop an oracle counterfactual framework for multi-turn problems that asks: how would an agent perform if it could leverage an oracle to perfectly perform a specific task? The change in the agent’s performance due to this oracle assistance allows us to measure the criticality of such oracle skill in the future advancement of AI agents. We introduce a suite of procedurally generated, game-like tasks with tunable complexity. These controlled environments allow us to provide precise oracle interventions, such as perfect planning or flawless state tracking, and make it possible to isolate the contribution of each oracle without confounding effects present in real-world benchmarks. Our results show that while some interventions (e.g., planning) consistently improve performance across settings, the usefulness of other skills is dependent on the properties of the environment and language model. Our work sheds light on the challenges of multi-turn agentic environments to guide the future efforts in the development of AI agents and language models.
arXiv:2601.16649v1 Announce Type: new
Abstract: Large language models can perform well on many isolated tasks, yet they continue to struggle on multi-turn, long-horizon agentic problems that require skills such as planning, state tracking, and long context processing. In this work, we aim to better understand the relative importance of advancing these underlying capabilities for success on such tasks. We develop an oracle counterfactual framework for multi-turn problems that asks: how would an agent perform if it could leverage an oracle to perfectly perform a specific task? The change in the agent’s performance due to this oracle assistance allows us to measure the criticality of such oracle skill in the future advancement of AI agents. We introduce a suite of procedurally generated, game-like tasks with tunable complexity. These controlled environments allow us to provide precise oracle interventions, such as perfect planning or flawless state tracking, and make it possible to isolate the contribution of each oracle without confounding effects present in real-world benchmarks. Our results show that while some interventions (e.g., planning) consistently improve performance across settings, the usefulness of other skills is dependent on the properties of the environment and language model. Our work sheds light on the challenges of multi-turn agentic environments to guide the future efforts in the development of AI agents and language models. Read More
Scalable Back-End for an AI-Based Diabetes Prediction Applicationcs.AI updates on arXiv.org arXiv:2512.08147v2 Announce Type: replace
Abstract: The rising global prevalence of diabetes necessitates early detection to prevent severe complications. While AI-powered prediction applications offer a promising solution, they require a responsive and scalable back-end architecture to serve a large user base effectively. This paper details the development and evaluation of a scalable back-end system designed for a mobile diabetes prediction application. The primary objective was to maintain a failure rate below 5% and an average latency of under 1000 ms. The architecture leverages horizontal scaling, database sharding, and asynchronous communication via a message queue. Performance evaluation showed that 83% of the system’s features (20 out of 24) met the specified performance targets. Key functionalities such as user profile management, activity tracking, and read-intensive prediction operations successfully achieved the desired performance. The system demonstrated the ability to handle up to 10,000 concurrent users without issues, validating its scalability. The implementation of asynchronous communication using RabbitMQ proved crucial in minimizing the error rate for computationally intensive prediction requests, ensuring system reliability by queuing requests and preventing data loss under heavy load.
arXiv:2512.08147v2 Announce Type: replace
Abstract: The rising global prevalence of diabetes necessitates early detection to prevent severe complications. While AI-powered prediction applications offer a promising solution, they require a responsive and scalable back-end architecture to serve a large user base effectively. This paper details the development and evaluation of a scalable back-end system designed for a mobile diabetes prediction application. The primary objective was to maintain a failure rate below 5% and an average latency of under 1000 ms. The architecture leverages horizontal scaling, database sharding, and asynchronous communication via a message queue. Performance evaluation showed that 83% of the system’s features (20 out of 24) met the specified performance targets. Key functionalities such as user profile management, activity tracking, and read-intensive prediction operations successfully achieved the desired performance. The system demonstrated the ability to handle up to 10,000 concurrent users without issues, validating its scalability. The implementation of asynchronous communication using RabbitMQ proved crucial in minimizing the error rate for computationally intensive prediction requests, ensuring system reliability by queuing requests and preventing data loss under heavy load. Read More
Cold snap highlight’s airlines’ proactive use of AIAI News The severe weather experienced at present in the US has placed significant strain on the airline industry in the country, with knock-on effects of changes to schedules and routes affecting the rest of the world. It’s at times like this that companies have to respond to queries from customers at a much greater rate than
The post Cold snap highlight’s airlines’ proactive use of AI appeared first on AI News.
The severe weather experienced at present in the US has placed significant strain on the airline industry in the country, with knock-on effects of changes to schedules and routes affecting the rest of the world. It’s at times like this that companies have to respond to queries from customers at a much greater rate than
The post Cold snap highlight’s airlines’ proactive use of AI appeared first on AI News. Read More
Lowering the barriers databases place in the way of strategy, with RavenDBAI News If database technologies offered performance, flexibility and security, most professionals would be happy to get two of the three, and they might have to expect to accept some compromises, too. Systems optimised for speed demand manual tuning, while flexible platforms can impose costs when early designs become constraints. Security is, sadly, sometimes, a bolt-on, with
The post Lowering the barriers databases place in the way of strategy, with RavenDB appeared first on AI News.
If database technologies offered performance, flexibility and security, most professionals would be happy to get two of the three, and they might have to expect to accept some compromises, too. Systems optimised for speed demand manual tuning, while flexible platforms can impose costs when early designs become constraints. Security is, sadly, sometimes, a bolt-on, with
The post Lowering the barriers databases place in the way of strategy, with RavenDB appeared first on AI News. Read More
Layered Architecture for Building Readable, Robust, and Extensible AppsTowards Data Science If adding a feature feels like open-heart surgery on your codebase, the problem isn’t bugs, it’s structure. This article shows how better architecture reduces risk, speeds up change, and keeps teams moving.
The post Layered Architecture for Building Readable, Robust, and Extensible Apps appeared first on Towards Data Science.
If adding a feature feels like open-heart surgery on your codebase, the problem isn’t bugs, it’s structure. This article shows how better architecture reduces risk, speeds up change, and keeps teams moving.
The post Layered Architecture for Building Readable, Robust, and Extensible Apps appeared first on Towards Data Science. Read More
Nonprofit security organization Shadowserver has found over 6,000 SmarterMail servers exposed online and likely vulnerable to attacks exploiting a critical authentication bypass vulnerability. […] Read More
