Agent Policy-Value Audit: Separating Transition Composition from Event Selection in Financial LLM AgentsResearchOct 6, 2026
Self-Propagating Misalignment in LLM Agents, and Why Auditing or Disabling Memory Is Not EnoughResearchOct 6, 2026
ClasSAE: Class-Aligned Sparse Autoencoders via Differentiable Feature-Class AffinityResearchOct 6, 2026
Can LLMs Separate Pasted Artifacts from User Speech? Absorption at Unmarked Prompt SeamsResearchOct 6, 2026
Conformal Prediction with Paraphrase-Aware Scoring for LLM Uncertainty QuantificationResearchOct 6, 2026