Adaptive Parallel MCTS for Fast Test-Time Compute Scaling
Boost large language model efficiency with adaptive parallel MCTS, reducing latency and improving throughput during test-time compute scaling.
The Silicon Mirror: Reducing Sycophancy in LLMs
Discover how The Silicon Mirror framework dynamically reduces sycophantic behavior in large language models, ensuring factual accuracy and trust.
Agent-Based AI Evaluation: Log Scores & Power-Law Insights
Explore how logarithmic scores and power-law patterns improve reliability in agent-based AI evaluations with persona-driven judges.
Reliable Truth-Aligned Uncertainty Estimation for LLMs
Enhance large language models' reliability with Truth AnChoring, a method for accurate truth-aligned uncertainty estimation and calibration.
Execution-Verified Reinforcement Learning for Optimization
Discover EVOM, a reinforcement learning framework enhancing optimization modeling with solver-specific code and zero-shot solver transfer.