PivotRL: High Accuracy Agentic Post-Training at Low Compute Cost

• URL: http://arxiv.org/abs/2603.21383v1
• Date: Sun, 22 Mar 2026 19:59:48 GMT
• Title: PivotRL: High Accuracy Agentic Post-Training at Low Compute Cost
• Abstract summary: Post-training for long-horizon agentic tasks has a tension between compute efficiency and generalization.<n>We introduce PivotRL, a novel framework that operates on existing SFT trajectories to combine the compute efficiency of SFT with the OOD accuracy of E2E RL.<n> PivotRL is adopted by NVIDIA's Nemotron-3-Super-120B-A12B, acting as the workhorse in production-scale agentic post-training.
• Score: 22.906887375657664
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Post-training for long-horizon agentic tasks has a tension between compute efficiency and generalization. While supervised fine-tuning (SFT) is compute efficient, it often suffers from out-of-domain (OOD) degradation. Conversely, end-to-end reinforcement learning (E2E RL) preserves OOD capabilities, but incurs high compute costs due to many turns of on-policy rollout. We introduce PivotRL, a novel framework that operates on existing SFT trajectories to combine the compute efficiency of SFT with the OOD accuracy of E2E RL. PivotRL relies on two key mechanisms: first, it executes local, on-policy rollouts and filters for pivots: informative intermediate turns where sampled actions exhibit high variance in outcomes; second, it utilizes rewards for functional-equivalent actions rather than demanding strict string matching with the SFT data demonstration. We theoretically show that these mechanisms incentivize strong learning signals with high natural gradient norm, while maximally preserving policy probability ordering on actions unrelated to training tasks. In comparison to standard SFT on identical data, we demonstrate that PivotRL achieves +4.17% higher in-domain accuracy on average across four agentic domains, and +10.04% higher OOD accuracy in non-agentic tasks. Notably, on agentic coding tasks, PivotRL achieves competitive accuracy with E2E RL with 4x fewer rollout turns. PivotRL is adopted by NVIDIA's Nemotron-3-Super-120B-A12B, acting as the workhorse in production-scale agentic post-training.

Related papers

• Unified Data Selection for LLM Reasoning [66.0118347918729]
  Training Large Language Models (LLMs) for complex, long-CoT reasoning is often bottlenecked by the need for massive high-quality reasoning data.<n>We propose High-Entropy Sum (HES), a training-free metric that quantifies reasoning quality by summing only the entropy of the top (e.g., 0.5%) highest-entropy tokens in each reasoning sample.
  arXiv  Detail & Related papers  (2026-05-21T12:21:41Z)
• From SFT to RL: Demystifying the Post-Training Pipeline for LLM-based Vulnerability Detection [2.6678231901651723]
  We present the first comprehensive investigation into the post-training pipeline for LLM-based vulnerability detection.<n>Our study identifies practical guidelines and insights, including how data curation, stage interactions, reward mechanisms, and evaluation protocols collectively dictate the efficacy of model training and assessment.
  arXiv  Detail & Related papers  (2026-02-15T06:33:25Z)
• CVeDRL: An Efficient Code Verifier via Difficulty-aware Reinforcement Learning [57.24524263804788]
  Code verifiers play a critical role in post-verification for LLM-based code generation.<n>Existing supervised fine-tuning methods suffer from data scarcity, high failure rates, and poor inference efficiency.<n>We show that naive RL with only functionality rewards fails to generate effective unit tests for difficult branches and samples.
  arXiv  Detail & Related papers  (2026-01-30T10:33:29Z)
• Trust-Region Adaptive Policy Optimization [82.09255251747818]
  Post-training methods play an important role in improving large language models' (LLMs) complex reasoning abilities.<n>We introduce TRAPO, a framework that interleavesSupervised Fine-Tuning (SFT) and Reinforcement Learning (RL) within each training instance.<n>Experiments on five mathematical reasoning benchmarks show that TRAPO consistently surpasses standard SFT, RL, and SFT-then-RL pipelines.
  arXiv  Detail & Related papers  (2025-12-19T14:37:07Z)
• Human-in-the-loop Online Rejection Sampling for Robotic Manipulation [55.99788088622936]
  Hi-ORS stabilizes value estimation by filtering out negatively rewarded samples during online fine-tuning.<n>Hi-ORS fine-tunes a pi-base policy to master contact-rich manipulation in just 1.5 hours of real-world training.
  arXiv  Detail & Related papers  (2025-10-30T11:53:08Z)
• Quagmires in SFT-RL Post-Training: When High SFT Scores Mislead and What to Use Instead [20.446287312285648]
  We study whether high SFT scores translate to improved performance after RL.<n>We find high SFT scores can be biased toward simpler or more homogeneous data and are not reliably predictive of subsequent RL gains or scaled-up post-training effectiveness.<n>We study alternative metrics and identify generalization loss on held-out reasoning examples and Pass@large k performance to provide strong proxies for the RL outcome.
  arXiv  Detail & Related papers  (2025-10-02T02:57:00Z)
• Beyond Two-Stage Training: Cooperative SFT and RL for LLM Reasoning [36.06085913761571]
  This study introduces a novel method for learning reasoning models that employs bilevel optimization.<n>By conditioning the SFT objective on the optimal RL policy, our approach enables SFT to meta-learn how to guide RL's optimization process.
  arXiv  Detail & Related papers  (2025-09-08T17:58:02Z)
• Your Reward Function for RL is Your Best PRM for Search: Unifying RL and Search-Based TTS [62.22644307952087]
  We introduce AIRL-S, the first natural unification of RL-based and search-based TTS.<n>We leverage adversarial inverse reinforcement learning (AIRL) combined with group relative policy optimization (GRPO) to learn a dense, dynamic PRM directly from correct reasoning traces.<n>Our results show that our unified approach improves performance by 9 % on average over the base model, matching GPT-4o.
  arXiv  Detail & Related papers  (2025-08-19T23:41:15Z)
• InfiAlign: A Scalable and Sample-Efficient Framework for Aligning LLMs to Enhance Reasoning Capabilities [27.09178257629886]
  InfiAlign is a scalable and sample-efficient post-training framework for large language models (LLMs)<n>At the core of InfiAlign is a robust data selection pipeline that automatically curates high-quality alignment data from open-source reasoning.<n>Our results highlight the effectiveness of combining principled data selection with full-stage post-training.
  arXiv  Detail & Related papers  (2025-08-07T15:34:06Z)
• How to Train Your LLM Web Agent: A Statistical Diagnosis [96.86317871461834]
  We present the first statistically grounded study on compute allocation for LLM web-agent post-training.<n>Our approach uses a two-stage pipeline, training a Llama 3.1 8B student to imitate a Llama 3.3 70B teacher via supervised fine-tuning (SFT) and on-policy reinforcement learning.<n>Our results show that combining SFT with on-policy RL consistently outperforms either approach alone on both WorkArena and MiniWob++.
  arXiv  Detail & Related papers  (2025-07-05T17:12:33Z)
• Learning to Optimize for Reinforcement Learning [58.01132862590378]
  Reinforcement learning (RL) is essentially different from supervised learning, and in practice, these learneds do not work well even in simple RL tasks. Agent-gradient distribution is non-independent and identically distributed, leading to inefficient meta-training. We show that, although only trained in toy tasks, our learned can generalize unseen complex tasks in Brax.
  arXiv  Detail & Related papers  (2023-02-03T00:11:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.