Johns Hopkins University · 15 records

dateidtitleauthorsvenue↑
2026-11-132611.12750The Narrative Gap: Can LLMs Help Us Navigate Diverse Narratives Across Languages?Nikhil Sharma et al.arXiv2
2026-11-122611.08538Subliminal Learning in Vision-Language ModelsArman Hatami et al.arXiv5
2026-11-02W7114777520Spatially Resolved Physical Properties of Young Star Clusters and Star-forming Clumps in the Brightest z > 6 Galaxy, the Strongly Lensed Cosmic Spear at z = 6.2Abdurrouf et al.The Astrophysical Journal
2026-10-092610.12360Accurate but Not Humble: Evaluating Epistemic Humility in LLM Agents under Knowledge ConflictKaiser Sun et al.arXiv10
2026-09-212609.22039Gricea: An Open Science Platform for Conversational AI ResearchNikhil Sharma et al.arXiv11
2026-09-162609.06986Continual Learning Mechanisms Compose for Long-Horizon MemorizationZheyuan Zhang et al.arXiv94
2026-09-112609.05903EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing AgentsNanxi Li et al.arXiv20
2026-07-272607.19636Multimodal Speaker Verification as a Threat to Speaker AnonymizationAshi Garg et al.arXiv1
2026-07-022606.28322PerceptionRubrics: Calibrating Multimodal Evaluation to Human PerceptionYana Wei et al.arXiv41
2026-06-292606.24893AgentOdyssey: Open-Ended Long-Horizon Text Game Generation for Test-Time Continual Learning AgentsZheyuan Zhang et al.arXiv7
2026-06-022606.00240MindZero: Learning Online Mental Reasoning With Zero AnnotationsShunchi Zhang et al.arXiv3
2026-05-252605.19354Next-Acceleration-Scale Prediction for Autoregressive MRI ReconstructionYilmaz Korkmaz et al.arXiv1
2026-05-202605.20087ThoughtTrace: Understanding User Thoughts in Real-World LLM InteractionsChuanyang Jin et al.arXiv16
2026-05-182604.09839Steered LLM Activations are Non-SurjectiveAayush Mishra et al.arXiv14
2026-05-132605.12449LychSim: A Controllable and Interactive Simulation Framework for Vision ResearchWufei Ma et al.arXiv3