SCOPE-AD: Sequential cost-aware ordinal-belief planning with energy-based models for diagnostic agentsResearchOct 2, 2026
When Does Exercise-Specific Joint Selection Help? An Audit of Evaluation and Control DesignResearchOct 2, 2026
Precision over Scale: A Polish-Silesian Benchmark and a Translation System Outperforming Open-Source and Commercial ModelsResearchOct 2, 2026
AgSpec: Pushing the Limits of Retrieval-Based Speculative Decoding in Coding Agent PipelinesResearchOct 2, 2026