The Continuity Layer: Why Intelligence Needs an Architecture for What It Carries Forward Paper • 2604.17273 • Published 3 days ago • 2
Training LLM Agents for Spontaneous, Reward-Free Self-Evolution via World Knowledge Exploration Paper • 2604.18131 • Published 2 days ago • 4 • 1
Extending One-Step Image Generation from Class Labels to Text via Discriminative Text Representation Paper • 2604.18168 • Published 2 days ago • 88 • 2
Revisiting a Pain in the Neck: A Semantic Reasoning Benchmark for Language Models Paper • 2604.16593 • Published 5 days ago • 3 • 2
Modeling Sparse and Bursty Vulnerability Sightings: Forecasting Under Data Constraints Paper • 2604.16038 • Published 5 days ago • 1 • 2
Concrete Jungle: Towards Concreteness Paved Contrastive Negative Mining for Compositional Understanding Paper • 2604.13313 • Published 8 days ago • 9 • 2
KWBench: Measuring Unprompted Problem Recognition in Knowledge Work Paper • 2604.15760 • Published 5 days ago • 2
Geometric coherence of single-cell CRISPR perturbations reveals regulatory architecture and predicts cellular stress Paper • 2604.16642 • Published 5 days ago • 1 • 2
EvoMaster: A Foundational Agent Framework for Building Evolving Autonomous Scientific Agents at Scale Paper • 2604.17406 • Published 3 days ago • 1 • 1
When Background Matters: Breaking Medical Vision Language Models by Transferable Attack Paper • 2604.17318 • Published 3 days ago • 1 • 2
GenericAgent: A Token-Efficient Self-Evolving LLM Agent via Contextual Information Density Maximization (V1.0) Paper • 2604.17091 • Published 4 days ago • 6 • 2
GFT: From Imitation to Reward Fine-Tuning with Unbiased Group Advantages and Dynamic Coefficient Rectification Paper • 2604.14258 • Published 7 days ago • 19 • 3
OneVL: One-Step Latent Reasoning and Planning with Vision-Language Explanation Paper • 2604.18486 • Published 2 days ago • 68 • 3
Beyond Text-Dominance: Understanding Modality Preference of Omni-modal Large Language Models Paper • 2604.16902 • Published 4 days ago • 2 • 2
The Geometric Canary: Predicting Steerability and Detecting Drift via Representational Stability Paper • 2604.17698 • Published 2 days ago • 1 • 2
The Illusion of Certainty: Decoupling Capability and Calibration in On-Policy Distillation Paper • 2604.16830 • Published 4 days ago • 10 • 2
Latent Preference Modeling for Cross-Session Personalized Tool Calling Paper • 2604.17886 • Published 2 days ago • 1 • 2