Author: Tech Jacks Solutions

HugAgent: Benchmarking LLMs for Simulation of Individualized Human Reasoningcs.AI updates on arXiv.org arXiv:2510.15144v3 Announce Type: replace
Abstract: Simulating human reasoning in open-ended tasks has long been a central aspiration in AI and cognitive science. While large language models now approximate human responses at scale, they remain tuned to population-level consensus, often erasing the individuality of reasoning styles and belief trajectories. To advance the vision of more human-like reasoning in machines, we introduce HugAgent (Human-Grounded Agent Benchmark), which rethinks human reasoning simulation along three dimensions: (i) from averaged to individualized reasoning, (ii) from behavioral mimicry to cognitive alignment, and (iii) from vignette-based to open-ended data. The benchmark evaluates whether a model can predict a specific person’s behavioral responses and the underlying reasoning dynamics in out-of-distribution scenarios, given partial evidence of their prior views. HugAgent adopts a dual-track design: a human track that automates and scales the think-aloud method to collect ecologically valid human reasoning data, and a synthetic track for further scalability and systematic stress testing. This architecture enables low-cost, extensible expansion to new tasks and populations. Experiments with state-of-the-art language models reveal persistent adaptation gaps, positioning HugAgent as the first extensible benchmark for aligning machine reasoning with the individuality of human thought. The benchmark, along with its complete data collection pipeline and companion chatbot, is open-sourced as HugAgent (https://anonymous.4open.science/r/HugAgent) and TraceYourThinking (https://anonymous.4open.science/r/trace-your-thinking).

 arXiv:2510.15144v3 Announce Type: replace
Abstract: Simulating human reasoning in open-ended tasks has long been a central aspiration in AI and cognitive science. While large language models now approximate human responses at scale, they remain tuned to population-level consensus, often erasing the individuality of reasoning styles and belief trajectories. To advance the vision of more human-like reasoning in machines, we introduce HugAgent (Human-Grounded Agent Benchmark), which rethinks human reasoning simulation along three dimensions: (i) from averaged to individualized reasoning, (ii) from behavioral mimicry to cognitive alignment, and (iii) from vignette-based to open-ended data. The benchmark evaluates whether a model can predict a specific person’s behavioral responses and the underlying reasoning dynamics in out-of-distribution scenarios, given partial evidence of their prior views. HugAgent adopts a dual-track design: a human track that automates and scales the think-aloud method to collect ecologically valid human reasoning data, and a synthetic track for further scalability and systematic stress testing. This architecture enables low-cost, extensible expansion to new tasks and populations. Experiments with state-of-the-art language models reveal persistent adaptation gaps, positioning HugAgent as the first extensible benchmark for aligning machine reasoning with the individuality of human thought. The benchmark, along with its complete data collection pipeline and companion chatbot, is open-sourced as HugAgent (https://anonymous.4open.science/r/HugAgent) and TraceYourThinking (https://anonymous.4open.science/r/trace-your-thinking). Read More  

TOBUGraph: Knowledge Graph-Based Retrieval for Enhanced LLM Performance Beyond RAGcs.AI updates on arXiv.org arXiv:2412.05447v3 Announce Type: replace-cross
Abstract: Retrieval-Augmented Generation (RAG) is one of the leading and most widely used techniques for enhancing LLM retrieval capabilities, but it still faces significant limitations in commercial use cases. RAG primarily relies on the query-chunk text-to-text similarity in the embedding space for retrieval and can fail to capture deeper semantic relationships across chunks, is highly sensitive to chunking strategies, and is prone to hallucinations. To address these challenges, we propose TOBUGraph, a graph-based retrieval framework that first constructs the knowledge graph from unstructured data dynamically and automatically. Using LLMs, TOBUGraph extracts structured knowledge and diverse relationships among data, going beyond RAG’s text-to-text similarity. Retrieval is achieved through graph traversal, leveraging the extracted relationships and structures to enhance retrieval accuracy, eliminating the need for chunking configurations while reducing hallucination. We demonstrate TOBUGraph’s effectiveness in TOBU, a real-world application in production for personal memory organization and retrieval. Our evaluation using real user data demonstrates that TOBUGraph outperforms multiple RAG implementations in both precision and recall, significantly improving user experience through improved retrieval accuracy.

 arXiv:2412.05447v3 Announce Type: replace-cross
Abstract: Retrieval-Augmented Generation (RAG) is one of the leading and most widely used techniques for enhancing LLM retrieval capabilities, but it still faces significant limitations in commercial use cases. RAG primarily relies on the query-chunk text-to-text similarity in the embedding space for retrieval and can fail to capture deeper semantic relationships across chunks, is highly sensitive to chunking strategies, and is prone to hallucinations. To address these challenges, we propose TOBUGraph, a graph-based retrieval framework that first constructs the knowledge graph from unstructured data dynamically and automatically. Using LLMs, TOBUGraph extracts structured knowledge and diverse relationships among data, going beyond RAG’s text-to-text similarity. Retrieval is achieved through graph traversal, leveraging the extracted relationships and structures to enhance retrieval accuracy, eliminating the need for chunking configurations while reducing hallucination. We demonstrate TOBUGraph’s effectiveness in TOBU, a real-world application in production for personal memory organization and retrieval. Our evaluation using real user data demonstrates that TOBUGraph outperforms multiple RAG implementations in both precision and recall, significantly improving user experience through improved retrieval accuracy. Read More  

Joint Verification and Refinement of Language Models for Safety-Constrained Planningcs.AI updates on arXiv.org arXiv:2410.14865v2 Announce Type: replace
Abstract: Large language models possess impressive capabilities in generating programs (e.g., Python) from natural language descriptions to execute robotic tasks. However, these generated programs often contain errors that violate externally given task specifications. Without an effective method to verify their correctness, the reliable deployment of language models in real-world systems is practically infeasible. We develop a method that converts generated robot programs into an automaton-based representation and verifies them against task-relevant safety specifications. We establish a theorem that any arbitrary combination of the verified programs will also satisfy the safety specifications. Hence, the method eliminates the need to verify complex programs composed of multiple simpler ones, reducing computation complexity. We then introduce an automated fine-tuning procedure that leverages verification outcomes for supervision. By applying the theorem, this procedure only requires training the model to generate safe sub-components, thereby improving training efficiency. Empirical results on robot applications show a 30 percent increase in the probability of generating specification-compliant programs, with training time reduced by half compared to fine-tuning on generating full programs.

 arXiv:2410.14865v2 Announce Type: replace
Abstract: Large language models possess impressive capabilities in generating programs (e.g., Python) from natural language descriptions to execute robotic tasks. However, these generated programs often contain errors that violate externally given task specifications. Without an effective method to verify their correctness, the reliable deployment of language models in real-world systems is practically infeasible. We develop a method that converts generated robot programs into an automaton-based representation and verifies them against task-relevant safety specifications. We establish a theorem that any arbitrary combination of the verified programs will also satisfy the safety specifications. Hence, the method eliminates the need to verify complex programs composed of multiple simpler ones, reducing computation complexity. We then introduce an automated fine-tuning procedure that leverages verification outcomes for supervision. By applying the theorem, this procedure only requires training the model to generate safe sub-components, thereby improving training efficiency. Empirical results on robot applications show a 30 percent increase in the probability of generating specification-compliant programs, with training time reduced by half compared to fine-tuning on generating full programs. Read More  

Mapping the AI Education Surge: Which States and Schools Are Leading the Pack in 2025KDnuggets The 2025 AI Degree Report is out, tracking how the AI talent pipeline is shifting, with analyses and data visualizations provide a roadmap to what’s happening on the ground.

 The 2025 AI Degree Report is out, tracking how the AI talent pipeline is shifting, with analyses and data visualizations provide a roadmap to what’s happening on the ground. Read More  

Why Storytelling With Data Matters for Business and Data AnalystsTowards Data Science Data is driving the future of business and here’s how you can be prepared for that future
The post Why Storytelling With Data Matters for Business and Data Analysts appeared first on Towards Data Science.

 Data is driving the future of business and here’s how you can be prepared for that future
The post Why Storytelling With Data Matters for Business and Data Analysts appeared first on Towards Data Science. Read More  

Does More Data Always Yield Better Performance?Towards Data Science Exploring and challenging the conventional wisdom of “more data → better performance” by experimenting with the interactions between sample size, attribute set, and model complexity.
The post Does More Data Always Yield Better Performance? appeared first on Towards Data Science.

 Exploring and challenging the conventional wisdom of “more data → better performance” by experimenting with the interactions between sample size, attribute set, and model complexity.
The post Does More Data Always Yield Better Performance? appeared first on Towards Data Science. Read More  

How Clario automates clinical research analysis using generative AI on AWSArtificial Intelligence In this post, we demonstrate how Clario has used Amazon Bedrock and other AWS services to build an AI-powered solution that automates and improves the analysis of COA interviews.

 In this post, we demonstrate how Clario has used Amazon Bedrock and other AWS services to build an AI-powered solution that automates and improves the analysis of COA interviews. Read More  

What Does the End of GIL Mean for Python?KDnuggets The GIL is finally being dismantled. The implications aren’t just technical; they’re cultural. This shift could redefine how we write, scale, and even think about Python in the modern era.

 The GIL is finally being dismantled. The implications aren’t just technical; they’re cultural. This shift could redefine how we write, scale, and even think about Python in the modern era. Read More  

7 Steps to Mastering Data Storytelling for Business ImpactKDnuggets This infographic distills a reliable workflow for turning analysis into decisions, helping you master data storytelling for business impact.

 This infographic distills a reliable workflow for turning analysis into decisions, helping you master data storytelling for business impact. Read More  

Deep learning models are vulnerable, but adversarial examples are even more vulnerablecs.AI updates on arXiv.org arXiv:2511.05073v1 Announce Type: cross
Abstract: Understanding intrinsic differences between adversarial examples and clean samples is key to enhancing DNN robustness and detection against adversarial attacks. This study first empirically finds that image-based adversarial examples are notably sensitive to occlusion. Controlled experiments on CIFAR-10 used nine canonical attacks (e.g., FGSM, PGD) to generate adversarial examples, paired with original samples for evaluation. We introduce Sliding Mask Confidence Entropy (SMCE) to quantify model confidence fluctuation under occlusion. Using 1800+ test images, SMCE calculations supported by Mask Entropy Field Maps and statistical distributions show adversarial examples have significantly higher confidence volatility under occlusion than originals. Based on this, we propose Sliding Window Mask-based Adversarial Example Detection (SWM-AED), which avoids catastrophic overfitting of conventional adversarial training. Evaluations across classifiers and attacks on CIFAR-10 demonstrate robust performance, with accuracy over 62% in most cases and up to 96.5%.

 arXiv:2511.05073v1 Announce Type: cross
Abstract: Understanding intrinsic differences between adversarial examples and clean samples is key to enhancing DNN robustness and detection against adversarial attacks. This study first empirically finds that image-based adversarial examples are notably sensitive to occlusion. Controlled experiments on CIFAR-10 used nine canonical attacks (e.g., FGSM, PGD) to generate adversarial examples, paired with original samples for evaluation. We introduce Sliding Mask Confidence Entropy (SMCE) to quantify model confidence fluctuation under occlusion. Using 1800+ test images, SMCE calculations supported by Mask Entropy Field Maps and statistical distributions show adversarial examples have significantly higher confidence volatility under occlusion than originals. Based on this, we propose Sliding Window Mask-based Adversarial Example Detection (SWM-AED), which avoids catastrophic overfitting of conventional adversarial training. Evaluations across classifiers and attacks on CIFAR-10 demonstrate robust performance, with accuracy over 62% in most cases and up to 96.5%. Read More