The rapid evolution of cyber threats has outpaced traditional detection methodologies, necessitating innovative approaches capable of addressing the adaptive and complex behaviors of modern adversaries. A novel framework was introduced, leveraging Temporal-Correlation Graphs to model the intricate relationships and temporal patterns inherent in malicious operations. The approach dynamically captured behavioral anomalies, offering a robust mechanism for distinguishing between benign and malicious activities in real-time scenarios. Extensive experiments demonstrated the framework's effectiveness across diverse ransomware families, with consistently high precision, recall, and overall detection accuracy. Comparative evaluations highlighted its better performance over traditional signature-based and heuristic methods, particularly in handling polymorphic and previously unseen ransomware variants. The architecture was designed with scalability and modularity in mind, ensuring compatibility with enterprise-scale environments while maintaining resource efficiency. Analysis of encryption speeds, anomaly patterns, and temporal correlations provided deeper insights into the operational strategies of ransomware, validating the framework's adaptability to evolving threats. The research contributes to advancing cybersecurity technologies by integrating dynamic graph analytics and machine learning for future innovations in threat detection. Results from this study underline the potential for transforming the way organizations detect and mitigate complex cyberattacks.

Sparse autoencoders (SAEs) are an increasingly popular technique for interpreting the internal representations of transformers. In this paper, we apply SAEs to 'interpret' random transformers, i.e., transformers where the parameters are sampled IID from a Gaussian rather than trained on text data. We find that random and trained transformers produce similarly interpretable SAE latents, and we confirm this finding quantitatively using an open-source auto-interpretability pipeline. Further, we find that SAE quality metrics are broadly similar for random and trained transformers. We find that these results hold across model sizes and layers. We discuss a number of number interesting questions that this work raises for the use of SAEs and auto-interpretability in the context of mechanistic interpretability.

In this work, we initiate the systemic study of decision-theoretic metrics in the design and analysis of algorithms with machine-learned predictions. We introduce approaches based on both deterministic measures such as distance-based evaluation, that help us quantify how close the algorithm is to an ideal solution, as well as stochastic measures that allow us to balance the trade-off between the algorithm's performance and the risk associated with the imperfect oracle. These approaches help us quantify the algorithmic performance across the entire spectrum of prediction error, unlike several previous works that focus on few, and often extreme values of the error. We apply these techniques to two well-known problems from resource allocation and online decision making, namely contract scheduling and 1-max search.

Time-sensitive machine learning benefits from Sequential Probability Ratio Test (SPRT), which provides an optimal stopping time for early classification of time series. However, in finite horizon scenarios, where input lengths are finite, determining the optimal stopping rule becomes computationally intensive due to the need for backward induction, limiting practical applicability. We thus introduce FIRMBOUND, an SPRT-based framework that efficiently estimates the solution to backward induction from training data, bridging the gap between optimal stopping theory and real-world deployment. It employs density ratio estimation and convex function learning to provide statistically consistent estimators for sufficient statistic and conditional expectation, both essential for solving backward induction; consequently, FIRMBOUND minimizes Bayes risk to reach optimality. Additionally, we present a faster alternative using Gaussian process regression, which significantly reduces training time while retaining low deployment overhead, albeit with potential compromise in statistical consistency. Experiments across independent and identically distributed (i.i.d.), non-i.i.d., binary, multiclass, synthetic, and real-world datasets show that FIRMBOUND achieves optimalities in the sense of Bayes risk and speed-accuracy tradeoff. Furthermore, it advances the tradeoff boundary toward optimality when possible and reduces decision-time variance, ensuring reliable decision-making. Code is publicly available at https://github.com/Akinori-F-Ebihara/FIRMBOUND

Entropy-based detection methodologies have gained significant attention due to their ability to analyze structural irregularities within executable files, particularly in the identification of malicious software employing advanced obfuscation techniques. The Entropy-Synchronized Neural Hashing (ESNH) framework introduces a novel approach that leverages entropy-driven hash representations to classify software binaries based on their underlying entropy characteristics. Through the synchronization of entropy profiles with neural network architectures, the model generates robust and unique hash values that maintain stability even when faced with polymorphic and metamorphic transformations. Comparative analysis against traditional detection approaches revealed superior performance in identifying novel threats, reducing false-positive rates, and achieving consistent classification across diverse ransomware families. The incorporation of a self-regulating hash convergence mechanism further ensured that entropy-synchronized hashes remained invariant across executions, minimizing classification inconsistencies that often arise due to dynamic modifications in ransomware payloads. Experimental results demonstrated high detection rates across contemporary ransomware strains, with the model exhibiting resilience against encryption-based evasion mechanisms, code injection strategies, and reflective loading techniques. Unlike conventional detection mechanisms that rely on static signatures and heuristic analysis, the proposed entropy-aware classification framework adapts to emerging threats through an inherent ability to capture entropy anomalies within executable structures. The findings reinforce the potential of entropy-based detection in addressing the limitations of traditional methodologies while enhancing detection robustness against obfuscation and adversarial evasion techniques.

As LLM agents gain a greater capacity to cause harm, AI developers might increasingly rely on control measures such as monitoring to justify that they are safe. We sketch how developers could construct a "control safety case", which is a structured argument that models are incapable of subverting control measures in order to cause unacceptable outcomes. As a case study, we sketch an argument that a hypothetical LLM agent deployed internally at an AI company won't exfiltrate sensitive information. The sketch relies on evidence from a "control evaluation,"' where a red team deliberately designs models to exfiltrate data in a proxy for the deployment environment. The safety case then hinges on several claims: (1) the red team adequately elicits model capabilities to exfiltrate data, (2) control measures remain at least as effective in deployment, and (3) developers conservatively extrapolate model performance to predict the probability of data exfiltration in deployment. This safety case sketch is a step toward more concrete arguments that can be used to show that a dangerously capable LLM agent is safe to deploy.

We present a modular pipeline that automates the generation of stealthy jailbreak prompts derived from high-level content policies, enhancing LLM content moderation. First, we address query inefficiency and jailbreak strength by developing Graph of Attacks with Pruning (GAP), a method that utilizes strategies from prior jailbreaks, resulting in 92% attack success rate on GPT-3.5 using only 54% of the queries of the prior algorithm. Second, we address the cold-start issue by automatically generating seed prompts from the high-level policy using LLMs. Finally, we demonstrate the utility of these generated jailbreak prompts of improving content moderation by fine-tuning PromptGuard, a model trained to detect jailbreaks, increasing its accuracy on the Toxic-Chat dataset from 5.1% to 93.89%.

We present COS(M+O)S, a System 2-inspired framework for open-ended plot development that systematically explores the vast space of possible story expansions, enabling a 3B-parameter language model to approach the plot quality of a 70B model on select short-story tasks. The method accomplishes this by combining Monte Carlo Tree Search (MCTS), guided by a step-level value model that rewards moderate surprisal (curiosity) while penalizing incoherence, and Odds Ratio Preference Optimization (ORPO) to fine-tune the policy on high-value plot expansions. This iterative reinforcement learning loop systematically explores multiple candidate plot branches, backpropagates quality signals, and adapts the policy for faster convergence, notably shifting the policy from puzzle-based Chain-of-Thought to more character-driven storytelling. In small-scale tests with short-story prompts, 67%-77% of participants favored COS(M+O)S's highest-rated expansions over lower-rated ones, suggesting that our learned value function aligns. GPT-4o ratings further show that COS(M+O)S surpasses naive single-pass decoding from Llama 3.2 3B by 0.59 SD, coming within 0.06 SD of Llama 3.1 70B (no significant difference, p=0.93). Pairwise comparisons with o1 place COS(M+O)S 1.5 SD above the 3B baseline and find no statistically significant gap from 70B. Nevertheless, absolute story quality remains modest, constrained by the small model's capacity and limited training data.

Early detection of COVID-19 is crucial for effective treatment and controlling its spread. This study proposes a novel hybrid deep learning model for detecting COVID-19 from CT scan images, designed to assist overburdened medical professionals. Our proposed model leverages the strengths of VGG16, DenseNet121, and MobileNetV2 to extract features, followed by Principal Component Analysis (PCA) for dimensionality reduction, after which the features are stacked and classified using a Support Vector Classifier (SVC). We conducted comparative analysis between the proposed hybrid model and individual pre-trained CNN models, using a dataset of 2,108 training images and 373 test images comprising both COVID-positive and non-COVID images. Our proposed hybrid model achieved an accuracy of 98.93%, outperforming the individual models in terms of precision, recall, F1 scores, and ROC curve performance.

Variable selection plays a crucial role in enhancing modeling effectiveness across diverse fields, addressing the challenges posed by high-dimensional datasets of correlated variables. This work introduces a novel approach namely Knockoff with over-parameterization (Knoop) to enhance Knockoff filters for variable selection. Specifically, Knoop first generates multiple knockoff variables for each original variable and integrates them with the original variables into an over-parameterized Ridgeless regression model. For each original variable, Knoop evaluates the coefficient distribution of its knockoffs and compares these with the original coefficients to conduct an anomaly-based significance test, ensuring robust variable selection. Extensive experiments demonstrate superior performance compared to existing methods in both simulation and real-world datasets. Knoop achieves a notably higher Area under the Curve (AUC) of the Receiver Operating Characteristic (ROC) Curve for effectively identifying relevant variables against the ground truth by controlled simulations, while showcasing enhanced predictive accuracy across diverse regression and classification tasks. The analytical results further backup our observations.