arXiv:2512.05794v2 Announce Type: replace-cross Abstract: Sparse autoencoders (SAEs) are a mechanistic interpretability technique that have been used to provide insight into learned concepts within large protein language models. Here, we employ TopK and Ordered SAEs to investigate autoregressive antibody language models, and steer their generation. We show that TopK SAEs can reveal biologically meaningful latent […]
Cross-Domain Offshore Wind Power Forecasting: Transfer Learning Through Meteorological Clusters
arXiv:2601.19674v2 Announce Type: replace-cross Abstract: Ambitious decarbonisation targets are rapidly increasing the commission of new offshore wind farms. For these newly commissioned plants to run, accurate power forecasts are needed from the onset. These allow grid stability, good reserve management and efficient energy trading. Despite machine learning models having strong performances, they tend to require […]
Causal Concept Graphs in LLM Latent Space for Stepwise Reasoning
arXiv:2603.10377v2 Announce Type: replace-cross Abstract: Sparse autoencoders can localize where concepts live in language models, but not how they interact during multi-step reasoning. We propose Causal Concept Graphs (CCG): a directed acyclic graph over sparse, interpretable latent features, where edges capture learned causal dependencies between concepts. We combine task-conditioned sparse autoencoders for concept discovery with […]
LLM+Graph@VLDB’2025 Workshop Summary
arXiv:2604.02861v2 Announce Type: replace-cross Abstract: The integration of large language models (LLMs) with graph-structured data has become a pivotal and fast evolving research frontier, drawing strong interest from both academia and industry. The 2nd LLM+Graph Workshop, co-located with the 51st International Conference on Very Large Data Bases (VLDB 2025) in London, focused on advancing algorithms […]
VLAA-GUI: Knowing When to Stop, Recover, and Search, A Modular Framework for GUI Automation
arXiv:2604.21375v2 Announce Type: replace-cross Abstract: Autonomous GUI agents face two fundamental challenges: early stopping, where agents prematurely declare success without verifiable evidence, and repetitive loops, where agents cycle through the same failing actions without recovery. We present VLAA-GUI, a modular GUI agentic framework built around three integrated components that guide the system on when to […]
Clarifying the conceptual dimensions of representation in neuroscience
arXiv:2403.14046v5 Announce Type: replace Abstract: Despite the centrality of the notion of representation in neuroscience, the field lacks a unified framework for the concepts used to characterize representation, leading to disparate use of both terminology and measures associated with it. To offer clarification, we propose a core set of conceptual dimensions that characterize representations in […]
How attention simplifies mental representations for planning
arXiv:2506.09520v2 Announce Type: replace Abstract: Human planning is efficient–it frugally deploys limited cognitive resources to accomplish difficult tasks–and flexible–adapting to novel problems and environments. Computational approaches suggest that people construct simplified mental representations of their environment, balancing the complexity of a task representation with its utility. These models imply a nested optimisation in which planning […]
When Models Outthink Their Safety: Unveiling and Mitigating Self-Jailbreak in Large Reasoning Models
arXiv:2510.21285v4 Announce Type: replace Abstract: Large Reasoning Models (LRMs) achieve strong performance on complex multi-step reasoning, yet they still exhibit severe safety failures such as harmful content generation. Existing methods often apply coarse-grained constraints over the entire reasoning trajectories, which can undermine reasoning capability while failing to address the root causes of unsafe behavior. In […]
Context-Sensitive Abstractions for Reinforcement Learning with Parameterized Actions
arXiv:2512.20831v2 Announce Type: replace 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 […]
From Multi-Agent to Single-Agent: When Is Skill Distillation Beneficial?
arXiv:2604.01608v3 Announce Type: replace Abstract: Multi-agent systems (MAS) tackle complex tasks by distributing expertise, though this often comes at the cost of heavy coordination overhead, context fragmentation, and brittle phase ordering. Distilling a MAS into a single-agent skill can bypass these costs, but this conversion lacks a principled answer for when and what to distill. […]
PSI: A Benchmark for Human Interpretation and Response in Traffic Interactions
arXiv:2112.02604v3 Announce Type: replace-cross Abstract: Accurately modeling pedestrian intention and understanding driver decision-making processes are critical for the development of safe and socially aware autonomous driving systems. We introduce PSI, a benchmark dataset that captures the dynamic evolution of pedestrian crossing intentions from the driver’s perspective, enriched with human textual explanations that reflect the reasoning […]
Fast, close, non-singular and property-preserving approximations of entropic measures
arXiv:2505.14234v2 Announce Type: replace-cross Abstract: Entropic measures like Shannon entropy (SE), its quantum mechanical analogue von Neumann entropy, and Kullback-Leibler divergence (KL) are key components in many tools used in physics, information theory, machine learning (ML) and quantum computing. Besides of the significant amounts of SE and KL computations required in these fields, the singularity […]