DiscoveryRSI: Recursively Learning to Discover Valuable Hypotheses

Researchers propose Hybrid Reasoning Agents that combine latent reasoning with text-based interaction, achieving 10.4% performance gains and 35-55% token reduction.

Researchers developed LLM agents that automate scientific reproducibility checks, successfully reproducing 77.4% of claims in medical studies and flagging 9.5% for expert review.

Researchers introduce MFlow, an adaptive parameterization for pixel diffusion transformers that learns to interpolate between x₀- and u-prediction, plus ConvRefiner to address high-frequency deficits.

Researchers discovered that fine-tuning LLMs can create 'behavioral couplings' where training for one behavior unintentionally elicits other behaviors, making misalignment difficult to trace.

Researchers demonstrate how task-aligned spectral queries in parameterized quantum circuits improve learning performance by modulating signal frequencies without adding new sample-dependent information.