Consolidating Rewarded Perturbations for LLM Post-Training
cs.CL, cs.LG
Submitted: 2026-05-29
Updated: 2026-11-06
Code: https://github.com/oooranz/CoRP
Terminology
Sources
- Proximal Policy Optimization Algorithms
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- OpenAI o1 System Card
- Kimi k1.5: Scaling Reinforcement Learning with LLMs
- Qwen3 Technical Report
- MiMo-V2-Flash Technical Report
- Spurious Rewards: Rethinking Training Signals in RLVR
- Reasoning-Finetuning Repurposes Latent Representations in Base Models
- Neural Thickets: Diverse Task Experts Are Dense Around Pretrained Weights
- Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning
- Evolution Strategies as a Scalable Alternative to Reinforcement Learning
- Qwen2.5 Technical Report
- Olmo 3
- The Llama 3 Herd of Models
- Training Verifiers to Solve Math Word Problems
- Program Synthesis with Large Language Models
- The Blessing of Dimensionality in LLM Fine-tuning: A Variance-Curvature Perspective
- Learning to Reason in 13 Parameters
- HybridFlow: A Flexible and Efficient RLHF Framework
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data