arXiv:2601.20894v1 Announce Type: new Abstract: Prompt-based continual learning methods effectively mitigate catastrophic forgetting. However, most existing methods assign a fixed set of prompts to each task, completely isolating knowledge across tasks and resulting in suboptimal parameter utilization. To address this, we consider the practical needs of continual learning and propose a prompt-sharing framework. This framework constructs a global prompt pool and introduces a task-aware gated routing mechanism that sparsely activates a subset of prompts to achieve dynamic decoupling […]

Ver mais

arXiv:2603.28796v1 Announce Type: new Abstract: Boolean satisfiability (SAT) problem, the first problem proven to be NP-complete, has become a fundamental challenge in computational complexity, with widespread applications in optimization and verification across many domains. Despite significant algorithmic advances over the past two decades, the performance of SAT solvers has improved at a limited pace. Notably, the 2025 competition winner shows only about a 2X improvement over the 2006 winner in SAT Competition performance after nearly 20 years of […]

Ver mais

arXiv:2603.00601v1 Announce Type: new Abstract: AI code agents excel at isolated tasks yet struggle with complex, multi-file software engineering requiring understanding of how dozens of modules relate. We hypothesize these failures stem from inability to construct, maintain, and update coherent architectural beliefs during codebase exploration. We introduce Theory of Code Space (ToCS), a benchmark that evaluates this capability by placing agents in procedurally generated codebases under partial observability, requiring them to build structured belief states over module dependencies, […]

Ver mais

Artificial intelligence continues to dominate corporate boardrooms and financial markets, with companies racing to secure computing power, build massive data centers, and develop increasingly sophisticated AI models. This excitement has set off a wave of unprecedented spending across the tech world, making people even more convinced that AI will define the next era of economic growth. Yet, amid the optimism, a different concern is beginning to emerge. The question now isn’t whether AI can change everything, it’s whether […]

Ver mais

arXiv:2603.28776v1 Announce Type: new Abstract: Learning to generate images with internally repeated and periodic structures poses a fundamental challenge for machine learning and computer vision models, which are typically optimised for local texture statistics and semantic realism rather than global structural consistency. This limitation is particularly pronounced in applications requiring strict control over repetition scale, spacing, and boundary coherence, such as microtopographical biomaterial surfaces. In this work, biomaterial design serves as a use case to study conditional generation […]

Ver mais

arXiv:2601.00021v1 Announce Type: new Abstract: We present a physical theory of intelligence grounded in irreversible information processing in systems constrained by conservation laws. An intelligent system is modelled as a coupled agent-environment process whose evolution transforms information into goal-directed work. To connect information to physical state, we introduce the Conservation-Congruent Encoding (CCE) framework, in which encodings correspond to metastable basins of attraction whose separability is enforced by conservation laws. Within this framework, intelligence is defined as the amount […]

Ver mais

A new study introduces the “self-evolution trilemma,” proving that AI systems cannot simultaneously remain autonomous, isolated from external input, and perfectly aligned with human values. Experiments in a closed AI community called Moltbook demonstrate how consensus bias, safety drift, collusion, and hidden behaviors emerge naturally under isolation. The paper argues that safety is not a patchable feature but a structural design choice — forcing builders of autonomous AI to decide which constraint they are willing to relax.

Ver mais

arXiv:2601.15473v1 Announce Type: new Abstract: Training modern deep learning models is increasingly constrained by GPU memory and compute limits. While Randomized Numerical Linear Algebra (RandNLA) offers proven techniques to compress these models, the lack of a unified, production-grade library prevents widely adopting these methods. We present Panther, a PyTorch-compatible library that consolidates established RandNLA algorithms into a single high-performance framework. Panther engineers efficient, drop-in replacements for standard components including sketched linear layers, 2D convolution, multi-head attention, and randomized […]

Ver mais