Verifier Noise Is a Temperature: Testing Tilt Predictions in Finite-Budget LLM-Guided Search
cs.LG, cs.AI
Submitted: 2026-11-06
Updated: 2026-11-06
Terminology
Sources
- GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning
- ROC-n-reroll: How verifier imperfection affects test-time scaling
- The Winner's Curse in LLM Self-Improvement Loops: Selection Noise, Lock-in, and Acceptance Rules
- SMCEvolve: Principled Scientific Discovery via Sequential Monte Carlo Evolution
- Catastrophic Goodhart: regularizing RLHF with KL divergence does not mitigate heavy-tailed reward misspecification
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- Evolution or Illusion? Rethinking Evaluation in LLM Evolutionary Search
- LLM Evolutionary Search Strategies Should be Compared by Tokens
- The Limits of Inference Scaling Through Resampling
- Soft Best-of-n Sampling for Model Alignment
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis