mlsanityData to 7 Oct 2026

Motif-2-12.7B-Reasoning: A Practitioner's Guide to RL Training Recipes

arXiv 2512.11463 ↗First posted 11 Dec 2025cs.AIcited by 1

On the map: LLM reasoning → Reinforcement learning for reasoning

Abstract (arXiv)

We introduce Motif-2-12.7B-Reasoning, a 12.7B parameter language model designed to bridge the gap between open-weight systems and proprietary frontier models in complex reasoning and long-context understanding. Addressing the common challenges of model collapse and training instability in reasoning adaptation, we propose a comprehensive, reproducible training recipe spanning system, data, and algorithmic optimizations. Our approach combines memory-efficient infrastructure for 64K-token contexts using hybrid parallelism and kernel-level optimizations with a two-stage Supervised Fine-Tuning (SFT) curriculum that mitigates distribution mismatch through verified, aligned synthetic data. Furthermore, we detail a robust Reinforcement Learning Fine-Tuning (RLFT) pipeline that stabilizes training via difficulty-aware data filtering and mixed-policy trajectory reuse. Empirical results demonstrate that Motif-2-12.7B-Reasoning achieves performance comparable to models with significantly larger parameter counts across mathematics, coding, and agentic benchmarks, offering the community a competitive open model and a practical blueprint for scaling reasoning capabilities under realistic compute constraints.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning. Builds on 13 of its 20 references that are in the map. First posted on arXiv 11 Dec 2025.

Builds on

13 of 20 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025Revisiting Reinforcement Learning for LLM Reasoning from A Cross-Domain Perspective17 Jun
  2. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scale18 Mar

Other references · 11

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  2. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 1,906
  3. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,258
  4. 2025Group Sequence Policy Optimizationcited by 788
  5. 2023DeepSpeed Ulysses: System Optimizations for Enabling Training of Extreme Long Sequence Transformer Modelscited by 239
  6. 2025AIMO-2 Winning Solution: Building State-of-the-Art Mathematical Reasoning Models with OpenMathReasoning datasetcited by 165
  7. 2024Liger Kernel: Efficient Triton Kernels for LLM Trainingcited by 123
  8. 2025OpenCodeReasoning: Advancing Data Distillation for Competitive Codingcited by 106
  9. 2025rStar-Coder: Scaling Competitive Code Reasoning with a Large-Scale Verified Datasetcited by 49
  10. 2025OpenCodeReasoning-II: A Simple Test Time Scaling Approach via Self-Critiquecited by 24
  11. 2025Motif 2 12.7B technical reportcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026On the Role of Reasoning Patterns in the Generalization Discrepancy of Long Chain-of-Thought Supervised Fine-Tuningcited by 4

Similar, not cited

Close in wording, no citation either way

  1. 2026Leveraging Instruction Tuning and Merging for Reasoning Model Adaptation
  2. 2025Reshaping Reasoning in LLMs: A Theoretical Analysis of RL Training Dynamics through Pattern Selection
  3. 2025RL Squeezes, SFT Expands: A Comparative Study of Reasoning LLMs
  4. 2025A Survey of Reinforcement Learning for Large Reasoning Models
  5. 2025Scaling Up RL: Unlocking Diverse Reasoning in LLMs via Prolonged Training
  6. 2025Reasoning Vectors: Transferring Chain-of-Thought Capabilities via Task Arithmetic
  7. 2026The Art of Efficient Reasoning: Data, Reward, and Optimization
  8. 2025Reasoning Pattern Matters: Learning to Reason without Human Rationales