arXiv:2512.20968v1 Announce Type: cross Abstract: Distributed attention is a fundamental problem for scaling context window for Large Language Models (LLMs). The state-of-the-art method, Ring-Attention, suffers from scalability limitations due to its excessive communication traffic. This paper proposes a new distributed attention algorithm, Mesh-Attention, by rethinking the design space of distributed attention with a new matrix-based […]
Reflection Pretraining Enables Token-Level Self-Correction in Biological Sequence Models
arXiv:2512.20954v1 Announce Type: cross Abstract: Chain-of-Thought (CoT) prompting has significantly advanced task-solving capabilities in natural language processing with large language models. Unlike standard prompting, CoT encourages the model to generate intermediate reasoning steps, non-answer tokens, that help guide the model toward more accurate final outputs. These intermediate steps enable more complex reasoning processes such as […]
A Blockchain-Monitored Agentic AI Architecture for Trusted Perception-Reasoning-Action Pipelines
arXiv:2512.20985v1 Announce Type: new Abstract: The application of agentic AI systems in autonomous decision-making is growing in the areas of healthcare, smart cities, digital forensics, and supply chain management. Even though these systems are flexible and offer real-time reasoning, they also raise concerns of trust and oversight, and integrity of the information and activities upon […]
Memory-Efficient Acceleration of Block Low-Rank Foundation Models on Resource Constrained GPUs
arXiv:2512.20861v1 Announce Type: cross Abstract: Recent advances in transformer-based foundation models have made them the default choice for many tasks, but their rapidly growing size makes fitting a full model on a single GPU increasingly difficult and their computational cost prohibitive. Block low-rank (BLR) compression techniques address this challenge by learning compact representations of weight […]
Context-Sensitive Abstractions for Reinforcement Learning with Parameterized Actions
arXiv:2512.20831v1 Announce Type: new Abstract: Real-world sequential decision-making often involves parameterized action spaces that require both, decisions regarding discrete actions and decisions about continuous action parameters governing how an action is executed. Existing approaches exhibit severe limitations in this setting — planning methods demand hand-crafted action models, and standard reinforcement learning (RL) algorithms are designed […]
AgentMath: Empowering Mathematical Reasoning for Large Language Models via Tool-Augmented Agent
arXiv:2512.20745v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) like o3 and DeepSeek-R1 have achieved remarkable progress in natural language reasoning with long chain-of-thought. However, they remain computationally inefficient and struggle with accuracy when solving problems requiring complex mathematical operations. In this work, we present AgentMath, an agent framework that seamlessly integrates language models’ reasoning […]
Computational optimisation of slow cooling profiles for the cryopreservation of cells in suspension
arXiv:2512.20805v1 Announce Type: new Abstract: The cryopreservation of biological materials is a highly complex process, as it involves numerous factors such as the cooling and thawing procedures, the administration of cryoprotective agents (CPAs), as well as the type and composition of cells. While theoretical work has yielded a better understanding of the processes occurring during […]
Eidoku: A Neuro-Symbolic Verification Gate for LLM Reasoning via Structural Constraint Satisfaction
arXiv:2512.20664v1 Announce Type: new Abstract: Large Language Models (LLMs) frequently produce hallucinated statements that are assigned high likelihood by the model itself, exposing a fundamental limitation of probability-based verification. This suggests that hallucination is often not a low-confidence phenomenon, but a failure of structural consistency. In this work, we reformulate the verification of LLM reasoning […]
From Pilots to Practices: A Scoping Review of GenAI-Enabled Personalization in Computer Science Education
arXiv:2512.20714v1 Announce Type: new Abstract: Generative AI enables personalized computer science education at scale, yet questions remain about whether such personalization supports or undermines learning. This scoping review synthesizes 32 studies (2023-2025) purposively sampled from 259 records to map personalization mechanisms and effectiveness signals in higher-education computer science contexts. We identify five application domains: intelligent […]
Intrinsic limits of timekeeping precision in gene regulatory cascades
arXiv:2512.20933v1 Announce Type: new Abstract: Multiple cellular processes are triggered when the concentration of a regulatory protein reaches a critical threshold. Previous analyses have characterized timing statistics for single-gene systems. However, many biological timers are based on cascades of genes that activate each other sequentially. Here, we develop an analytical framework to describe the timing […]