Branch-and-Commit for Frozen RL Agents: When and Where to Spend Test-Time Compute
cs.LG
Submitted: 2026-11-08
Updated: 2026-11-08
Terminology
Sources
- Large Language Monkeys: Scaling Inference Compute with Repeated Sampling
- Training Verifiers to Solve Math Word Problems
- Guiding Pretraining in Reinforcement Learning with Large Language Models
- Reasoning with Language Model is Planning with World Model
- Agentic Test-Time Scaling for WebAgents
- Predicting and improving test-time scaling laws via reward tail-guided search
- What should post-training optimize? A test-time scaling law perspective
- Entropy-Gated Branching for Efficient Test-Time Reasoning
- Let's Verify Step by Step
- Eureka: Human-Level Reward Design via Coding Large Language Models
- Adaptive Inference-Time Compute: LLMs Can Predict if They Can Do Better, Even Mid-Generation
- Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learning
- WebGPT: Browser-assisted question-answering with human feedback
- XLand-MiniGrid: Scalable Meta-Reinforcement Learning Environments in JAX
- Assessing the Zero-Shot Capabilities of LLMs for Action Evaluation in RL
- Adaptive Test-Time Compute Allocation via Learned Heuristics over Categorical Structure
- Proximal Policy Optimization Algorithms
- Scaling Test-Time Compute Without Verification or RL is Suboptimal
- The Limits of Inference Scaling Through Resampling
- Voyager: An Open-Ended Embodied Agent with Large Language Models
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