Inspired by the success of large language models, there is a trend toward developing graph foundation models to conduct diverse downstream tasks in various domains. However, current models often require extra fine-tuning to apply their learned structural and semantic representations to new graphs, which limits their versatility.

This work presents the first theoretically justified simultaneous inference framework for off-policy evaluation (OPE). In contrast to existing methods that focus on point estimates or pointwise confidence intervals (CIs), the new framework quantifies global uncertainty across an infinite or continuous initial state space, offering valid inference over the entire state space.

In reinforcement learning with image-based inputs, it is crucial to establish a robust and generalizable state representation. Recent advancements in metric learning, such as deep bisimulation metric approaches, have shown promising results in learning structured low-dimensional representation space from pixel observations, where the distance between states is measured based on task-relevant features. However, these approaches face challenges in demanding generalization tasks and scenarios with non-informative rewards. This is because they fail to capture sufficient long-term information in the learned representations. To address these challenges, we propose a novel State Chrono Representation (SCR) approach. SCR augments state metric-based representations by incorporating extensive temporal information into the update step of bisimulation metric learning. It learns state distances within a temporal framework that considers both future dynamics and cumulative rewards over current and long-term future states. Our learning strategy effectively incorporates future behavioral information into the representation space without introducing a significant number of additional parameters for modeling dynamics. Extensive experiments conducted in DeepMind Control and Meta-World environments demonstrate that SCR achieves better performance comparing to other recent metric-based methods in demanding generalization tasks.

Large Language Models (LLMs) are increasingly deployed as autonomous agents, yet their practical utility is fundamentally constrained by a limited context window and state desynchronization resulting from the LLMs' stateless nature and inefficient context management. These limitations lead to unreliable output, unpredictable behavior, and inefficient resource usage, particularly when interacting with large, structured, and sensitive knowledge systems such as codebases and documents. To address these challenges, we introduce the Gatekeeper Protocol, a novel, domain-agnostic framework that governs agent-system interactions. Our protocol mandates that the agent first operate and reason on a minimalist, low-fidelity "latent state" representation of the system to strategically request high-fidelity context on demand. All interactions are mediated through a unified JSON format that serves as a declarative, state-synchronized protocol, ensuring the agent's model of the system remains verifiably grounded in the system's reality. We demonstrate the efficacy of this protocol with Sage, a reference implementation of the Gatekeeper Protocol for software development. Our results show that this approach significantly increases agent reliability, improves computational efficiency by minimizing token consumption, and enables scalable interaction with complex systems, creating a foundational methodology for building more robust, predictable, and grounded AI agents for any structured knowledge domain.

Pasur is a fishing card game played over six rounds and is played similarly to games such as Cassino and Scopa, and Bastra. This paper introduces a CUDA-accelerated computational framework for simulating Pasur, emphasizing efficient memory management. We use our framework to compute near-Nash equilibria via Counterfactual Regret Minimization (CFR), a well-known algorithm for solving large imperfect-information games. Solving Pasur presents unique challenges due to its intricate rules and the large size of its game tree. We handle rule complexity using PyTorch CUDA tensors and to address the memory-intensive nature of the game, we decompose the game tree into two key components: (1) actual game states, and (2) inherited scores from previous rounds. We construct the Full Game Tree by pairing card states with accumulated scores in the Unfolding Process. This design reduces memory overhead by storing only essential strategy values and node connections. To further manage computational complexity, we apply a round-by-round backward training strategy, starting from the final round and recursively propagating average utilities to earlier stages. Our approach constructs the complete game tree, which on average consists of over $10^9$ nodes. We provide detailed implementation snippets. After computing a near-Nash equilibrium strategy, we train a tree-based model to predict these strategies for use during gameplay. We then estimate the fair value of each deck through large-scale self-play between equilibrium strategies by simulating, for instance, 10,000 games per matchup, executed in parallel using GPU acceleration. Similar frameworks can be extended to other reinforcement learning algorithms where the action tree naturally decomposes into multiple rounds such as turn-based strategy games or sequential trading decisions in financial markets.

Decomposing Electrodermal Activity (EDA) into phasic (short-term, stimulus-linked responses) and tonic (longer-term baseline) components is essential for extracting meaningful emotional and physiological biomarkers. This study presents a comparative analysis of knowledge-driven, statistical, and deep learning-based methods for EDA signal decomposition, with a focus on in-the-wild data collected from wearable devices. In particular, the authors introduce the Feel Transformer, a novel Transformer-based model adapted from the Autoformer architecture, designed to separate phasic and tonic components without explicit supervision. The model leverages pooling and trend-removal mechanisms to enforce physiologically meaningful decompositions. Comparative experiments against methods such as Ledalab, cvxEDA, and conventional detrending show that the Feel Transformer achieves a balance between feature fidelity (SCR frequency, amplitude, and tonic slope) and robustness to noisy, real-world data. The model demonstrates potential for real-time biosignal analysis and future applications in stress prediction, digital mental health interventions, and physiological forecasting.

In reinforcement learning with image-based inputs, it is crucial to establish a robust and generalizable state representation. Recent advancements in metric learning, such as deep bisimulation metric approaches, have shown promising results in learning structured low-dimensional representation space from pixel observations, where the distance between states is measured based on task-relevant features. However, these approaches face challenges in demanding generalization tasks and scenarios with non-informative rewards. This is because they fail to capture sufficient long-term information in the learned representations. To address these challenges, we propose a novel State Chrono Representation (SCR) approach.

Large Language Models (LLMs) are known to be vulnerable to jailbreaking attacks, wherein adversaries exploit carefully engineered prompts to induce harmful or unethical responses. Such threats have raised critical concerns about the safety and reliability of LLMs in real-world deployment. While existing defense mechanisms partially mitigate such risks, subsequent advancements in adversarial techniques have enabled novel jailbreaking methods to circumvent these protections, exposing the limitations of static defense frameworks. In this work, we explore defending against evolving jailbreaking threats through the lens of context retrieval. First, we conduct a preliminary study demonstrating that even a minimal set of safety-aligned examples against a particular jailbreak can significantly enhance robustness against this attack pattern. Building on this insight, we further leverage the retrieval-augmented generation (RAG) techniques and propose Safety Context Retrieval (SCR), a scalable and robust safeguarding paradigm for LLMs against jailbreaking. Our comprehensive experiments demonstrate how SCR achieves superior defensive performance against both established and emerging jailbreaking tactics, contributing a new paradigm to LLM safety. Our code will be available upon publication.

Extended sequence generation often leads to degradation in contextual consistency due to the inability of conventional self-attention mechanisms to effectively retain long-range dependencies. Existing approaches, including memory compression and retrieval-augmented conditioning, introduce computational trade-offs that either increase inference latency or impose additional storage overhead. Structured Context Recomposition (SCR) introduces a probabilistic layer realignment strategy that dynamically adjusts learned representations within transformer layers, ensuring that semantically relevant embeddings persist throughout extended transformations. The proposed method enhances coherence retention through a recursive weighting function that redistributes representational emphasis based on inferred contextual relevance rather than relying on fixed token-level attention scores. Empirical results indicate that probabilistic realignment mitigates abrupt topic shifts and logical inconsistencies, particularly in scenarios where sequences exceed standard attention window constraints. Sequence-level entropy analysis further reveals that SCR moderates representational variability without introducing excessive output regularization, allowing models to sustain generative diversity while preserving contextual alignment. Attention head deviation measurements confirm that hierarchical reweighting contributes to smoother token dependency transitions across transformer layers, reinforcing the stability of multi-turn interactions and document-level reasoning. Computational resource assessments show that while SCR incurs a moderate increase in processing time, memory overhead remains within feasible limits, making it suitable for practical deployment in autoregressive generative applications.

An SCR that is assessed as relevant to safety goes through extra scrutiny to maintain personnel safety at the nuclear station. The current method of SCR classification is a manual one that involves human evaluators to examine multiple SCRs every week. These records, which may be submitted by any employee, cover a broad spectrum of events and undergo management review to determine an appropriate reaction. If an SCR is deemed relevant to safety, it undergoes further examination by the Health and Safety department and is documented in a specialized database. The SCR database encompasses a range of occurrences, from equipment malfunctions and delays in material delivery to staff missing training sessions, making it cumbersome for the Health and Safety department to sift through each SCR to identify safety-related items before transferring pertinent details into their safety tracking system. The aim of this project is to develop a machine learning classifier to automatically differentiate between safety-related and non-safety-related SCRs. While this tool is not intended to supplant human assessment, it will serve as an additional layer of scrutiny and facilitate the swift review of safetyrelated SCRs by triggering a pipeline that copies all relevant data into the safety system for final human verification.