# On the Interplay of Pre-Training, Mid-Training, and RL on Reasoning Language Models

arXiv: https://arxiv.org/abs/2512.07783 · HTML: https://mlsanity.com/paper/2512.07783/

## Key facts

- Cited by 69 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning.
- Builds on 20 of its 22 references that are in the map.
- First posted on arXiv 8 Dec 2025.

## Placed in

- LLM reasoning / [Reinforcement learning for reasoning](https://mlsanity.com/lr/reinforcement-learning-for-reasoning/)

## Builds on (most cited first, 20 of 20)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?](https://mlsanity.com/paper/2504.13837/) (2025)
- [GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models](https://mlsanity.com/paper/2410.05229/) (2024)
- [Process Reinforcement through Implicit Rewards](https://mlsanity.com/paper/2502.01456/) (2025)
- [Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRs](https://mlsanity.com/paper/2503.01307/) (2025)
- [Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains](https://mlsanity.com/paper/2507.17746/) (2025)
- [Reinforcement Learning with Verifiable Rewards Implicitly Incentivizes Correct Reasoning in Base LLMs](https://mlsanity.com/paper/2506.14245/) (2025)
- [Spurious Rewards: Rethinking Training Signals in RLVR](https://mlsanity.com/paper/2506.10947/) (2025)
- [ProRL: Prolonged Reinforcement Learning Expands Reasoning Boundaries in Large Language Models](https://mlsanity.com/paper/2505.24864/) (2025)
- [Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining](https://mlsanity.com/paper/2504.07912/) (2025)
- [Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process](https://mlsanity.com/paper/2407.20311/) (2024)
- [Does Math Reasoning Improve General LLM Capabilities? Understanding Transferability of LLM Reasoning](https://mlsanity.com/paper/2507.00432/) (2025)
- [The Invisible Leash: Why RLVR May or May Not Escape Its Origin](https://mlsanity.com/paper/2507.14843/) (2025)
- [e3: Learning to Explore Enables Extrapolation of Test-Time Compute for LLMs](https://mlsanity.com/paper/2506.09026/) (2025)
- [OMEGA: Can LLMs Reason Outside the Box in Math? Evaluating Exploratory, Compositional, and Transformative Generalization](https://mlsanity.com/paper/2506.18880/) (2025)
- [From f(x) and g(x) to f(g(x)): LLMs Learn New Skills in RL by Composing Old Ones](https://mlsanity.com/paper/2509.25123/) (2025)
- [Front-Loading Reasoning: The Synergy between Pretraining and Post-Training Data](https://mlsanity.com/paper/2510.03264/) (2025)
- [RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?](https://mlsanity.com/paper/2509.21016/) (2025)
- [Does Learning Mathematical Problem-Solving Generalize to Broader Reasoning?](https://mlsanity.com/paper/2507.04391/) (2025)

## Built on by (most cited first, 30 of 69)

- [Maximum Likelihood Reinforcement Learning](https://mlsanity.com/paper/2602.02710/) (2026)
- [Reinforcing Chain-of-Thought Reasoning with Self-Evolving Rubrics](https://mlsanity.com/paper/2602.10885/) (2026)
- [The Flexibility Trap: Rethinking the Value of Arbitrary Order in Diffusion Language Models](https://mlsanity.com/paper/2601.15165/) (2026)
- [daVinci-Dev: Agent-native Mid-training for Software Engineering](https://mlsanity.com/paper/2601.18418/) (2026)
- [GUI-GENESIS: Automated Synthesis of Efficient Environments with Verifiable Rewards for GUI Agent Post-Training](https://mlsanity.com/paper/2602.14093/) (2026)
- [Rebellious Student: Reversing Teacher Signals for Reasoning Exploration with Self-Distilled RLVR](https://mlsanity.com/paper/2605.10781/) (2026)
- [R3L: Reflect-then-Retry Reinforcement Learning with Language-Guided Exploration, Pivotal Credit, and Positive Amplification](https://mlsanity.com/paper/2601.03715/) (2026)
- [Reasoning Cache: Continual Improvement Over Long Horizons via Short-Horizon RL](https://mlsanity.com/paper/2602.03773/) (2026)
- [Learn Hard Problems During RL with Reference Guided Fine-tuning](https://mlsanity.com/paper/2603.01223/) (2026)
- [On-Policy Self-Distillation with Sampled Demonstrations Reduces Output Diversity](https://mlsanity.com/paper/2606.26091/) (2026)
- [Learning to Simulate Human Dialogue](https://mlsanity.com/paper/2601.04436/) (2026)
- [SpotAgent: Grounding Visual Geo-localization in Large Vision-Language Models through Agentic Reasoning](https://mlsanity.com/paper/2602.09463/) (2026)
- [Efficient Context Scaling with LongCat ZigZag Attention](https://mlsanity.com/paper/2512.23966/) (2025)
- [Outcome-Based RL Provably Leads Transformers to Reason, but Only With the Right Data](https://mlsanity.com/paper/2601.15158/) (2026)
- [Towards a Mechanistic Understanding of Large Reasoning Models: A Survey of Training, Inference, and Failures](https://mlsanity.com/paper/2601.19928/) (2026)
- [Epistemic Context Learning: Building Trust the Right Way in LLM-Based Multi-Agent Systems](https://mlsanity.com/paper/2601.21742/) (2026)
- [Autonomous Continual Learning for Environment Adaptation of Computer-Use Agents](https://mlsanity.com/paper/2602.10356/) (2026)
- [Aletheia: What Makes RLVR For Code Verifiers Tick?](https://mlsanity.com/paper/2601.12186/) (2026)
- [ASTER: Agentic Scaling with Tool-integrated Extended Reasoning](https://mlsanity.com/paper/2602.01204/) (2026)
- [Emergent Analogical Reasoning in Transformers](https://mlsanity.com/paper/2602.01992/) (2026)
- [UI-Oceanus: Scaling GUI Agents with Synthetic Environmental Dynamics](https://mlsanity.com/paper/2604.02345/) (2026)
- [StaRPO: Stability-Augmented Reinforcement Policy Optimization](https://mlsanity.com/paper/2604.08905/) (2026)
- [From Reasoning Traces to Reusable Modules: Understanding Compositional Generalization in Language Model Reasoning](https://mlsanity.com/paper/2606.18089/) (2026)
- [Understanding Reasoning from Pretraining to Post-Training](https://mlsanity.com/paper/2607.16097/) (2026)
- [A Relative-Budget Theory for Reinforcement Learning with Verifiable Rewards in Large Language Model Reasoning](https://mlsanity.com/paper/2602.01523/) (2026)
- [On the Role of Reasoning Patterns in the Generalization Discrepancy of Long Chain-of-Thought Supervised Fine-Tuning](https://mlsanity.com/paper/2604.01702/) (2026)
- [Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs](https://mlsanity.com/paper/2605.07153/) (2026)
- [VeriGate: Verifier-Gated Step-Level Supervision for GRPO](https://mlsanity.com/paper/2605.30451/) (2026)
- [SmartSnap: Proactive Evidence Seeking for Self-Verifying Agents](https://mlsanity.com/paper/2512.22322/) (2025)
- [Effective Reasoning Chains Reduce Intrinsic Dimensionality](https://mlsanity.com/paper/2602.09276/) (2026)

## Abstract (arXiv)

Recent reinforcement learning (RL) techniques have yielded impressive reasoning improvements in language models, yet it remains unclear whether post-training truly extends a model's reasoning ability beyond what it acquires during pre-training. A central challenge is the lack of control in modern training pipelines: large-scale pre-training corpora are opaque, mid-training is often underexamined, and RL objectives interact with unknown prior knowledge in complex ways. To resolve this ambiguity, we develop a fully controlled experimental framework that isolates the causal contributions of pre-training, mid-training, and RL-based post-training. Our approach employs synthetic reasoning tasks with explicit atomic operations, parseable step-by-step reasoning traces, and systematic manipulation of training distributions. We evaluate models along two axes: extrapolative generalization to more complex compositions and contextual generalization across surface contexts. Using this framework, we reconcile competing views on RL's effectiveness. We show that: 1) RL produces true capability gains (pass@128) only when pre-training leaves sufficient headroom and when RL data target the model's edge of competence, tasks at the boundary that are difficult but not yet out of reach. 2) Contextual generalization requires minimal yet sufficient pre-training exposure, after which RL can reliably transfer. 3) Mid-training significantly enhances performance under fixed compute compared with RL only, demonstrating its central but underexplored role in training pipelines. 4) Process-level rewards reduce reward hacking and improve reasoning fidelity. Together, these results clarify the interplay between pre-training, mid-training, and RL, offering a foundation for understanding and improving reasoning LM training strategies.