Primal Generation, Dual Judgment: Self-Training from Test-Time Scaling

• URL: http://arxiv.org/abs/2605.11299v2
• Date: Wed, 13 May 2026 03:29:35 GMT
• Title: Primal Generation, Dual Judgment: Self-Training from Test-Time Scaling
• Abstract summary: We introduce DuST (Dual Self-Training), a framework for self-training from the dual judgment space.<n>We show that DuST consistently improves Best-of-4 test-time scaling on LiveCodeBench.<n>For Qwen3-30B-Thinking on LiveCodeBench v6, judgment quality improves by +6.2 NDCG, single-sample pass@1 improves by +3.1, and Best-of-4 accuracy improves by +4.1.
• Score: 29.008352630270192
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Code generation is typically trained in the primal space of programs: a model produces a candidate solution and receives sparse execution feedback, often a single pass/fail bit. Test-time scaling enriches the inference procedure by sampling multiple candidates and judging among them, but the comparative information this process reveals is discarded after inference. We argue that this information defines a dual judgment space that provides a far richer training signal: the model learns not from an isolated success or failure, but from the relative correctness structure across its own plausible attempts, identifying which succeed, which fail, and what distinguishes them. We introduce DuST (Dual Self-Training), a framework for self-training from the dual judgment space. DuST samples candidate programs from the model's own distribution, labels them through sandbox execution, retains groups containing both successes and failures, and trains the model to rank candidates by execution correctness using GRPO. The objective is purely discriminative: the model is never directly rewarded for generating correct programs. Dual self-training improves both judgment and generation. Across five models spanning two families and three scales (4B to 30B), DuST consistently improves Best-of-4 test-time scaling on LiveCodeBench. For Qwen3-30B-Thinking on LiveCodeBench v6, judgment quality improves by +6.2 NDCG, single-sample pass@1 improves by +3.1, and Best-of-4 accuracy improves by +4.1. The trained model's single rollout matches the base model's Best-of-4 performance. SFT on the same ranking data improves judgment without improving generation, confirming that on-policy RL is the mechanism that transfers dual-space learning back into primal generation.

Related papers

• REVES: REvision and VErification--Augmented Training for Test-Time Scaling [53.197756110943395]
  We propose a two-stage iterative framework that alternates between online data/prompt augmentation and policy optimization.<n>We observe gains of +6.5 points over the RL baseline and +4.0 points over standard multi-turn training.
  arXiv  Detail & Related papers  (2026-06-17T10:37:23Z)
• FABSVer: Faster Training and Better Self-Verification for LLM Mathematical Reasoning [8.96133612398978]
  Existing approaches that equip models with self-verification typically treat solution generation and verification as two separate tasks.<n>We propose FABSVer, which fuses these two tasks into a single generation pass.<n>We show that FABSVer achieves superior self-verification and reasoning performance across three model scales.
  arXiv  Detail & Related papers  (2026-05-27T12:26:34Z)
• Self-Verified Distillation: Your Language Model Is Secretly Its Own Synthetic Data Pipeline [56.53954182896384]
  We propose Self-Verified Distillation, a simple post-training refinement algorithm for large language models.<n>Self-Verified Distillation generates candidate solutions to unlabeled seed questions.<n>It filters them using prompt-based self-verification, and trains on the resulting self-curated dataset.<n>We find that sampling more candidate generations and using a larger verification budget during training data construction produces higher-quality self-curated data.
  arXiv  Detail & Related papers  (2026-05-20T17:26:10Z)
• JudgeRLVR: Judge First, Generate Second for Efficient Reasoning [20.448286296459344]
  Reinforcement Learning with Verifiable Rewards (RLVR) has become a standard paradigm for reasoning in Large Language Models.<n>In this paper, we argue that discriminative capability is a prerequisite for efficient generation.<n>We propose JudgeRLVR, a two-stage judge-then-generate paradigm.
  arXiv  Detail & Related papers  (2026-01-13T11:47:42Z)
• Self-Improving VLM Judges Without Human Annotations [74.29324865147838]
  We present a framework to self-train a VLM judge model without any human preference annotations, using only self-synthesized data.<n>Our method improves a Llama-3.2-11B multimodal judge from 0.38 to 0.51 in overall accuracy on Multimodal RewardBench.<n>The overall strength of these human-annotation-free results suggest the potential for a future self-judge that evolves alongside rapidly improving VLM capabilities.
  arXiv  Detail & Related papers  (2025-12-02T20:52:19Z)
• Continuous Self-Improvement of Large Language Models by Test-time Training with Verifier-Driven Sample Selection [6.471199527741301]
  We introduce a new framework called VDS-TTT - Verifier-Driven Sample Selection for Test-Time Training.<n>We use a learned verifier to score a pool of generated responses and select only from high ranking pseudo-labeled examples for fine-tuned adaptation.<n>We fine-tune only low-rank LoRA adapter parameters, ensuring adaptation efficiency and fast convergence.
  arXiv  Detail & Related papers  (2025-05-26T03:54:47Z)
• Learning to Solve and Verify: A Self-Play Framework for Code and Test Generation [69.62857948698436]
  Recent advances in large language models (LLMs) have improved their performance on coding benchmarks.<n>However, improvement is plateauing due to the exhaustion of readily available high-quality data.<n>We propose Sol-Ver, a self-play solver-verifier framework that jointly improves a single model's code and test generation capacity.
  arXiv  Detail & Related papers  (2025-02-20T18:32:19Z)
• S$^2$R: Teaching LLMs to Self-verify and Self-correct via Reinforcement Learning [51.84977135926156]
  We introduce S$2$R, an efficient framework that enhances LLM reasoning by teaching models to self-verify and self-correct during inference.<n>Our results demonstrate that Qwen2.5-math-7B achieves an accuracy improvement from 51.0% to 81.6%, outperforming models trained on an equivalent amount of long-CoT distilled data.
  arXiv  Detail & Related papers  (2025-02-18T13:40:22Z)
• Training Language Models to Self-Correct via Reinforcement Learning [98.35197671595343]
  Self-correction has been found to be largely ineffective in modern large language models (LLMs) We develop a multi-turn online reinforcement learning approach, SCoRe, that significantly improves an LLM's self-correction ability using entirely self-generated data. We find that SCoRe achieves state-of-the-art self-correction performance, improving the base models' self-correction by 15.6% and 9.1% respectively on MATH and HumanEval.
  arXiv  Detail & Related papers  (2024-09-19T17:16:21Z)
• The Right Tool for the Job: Matching Model and Instance Complexities [62.95183777679024]
  As NLP models become larger, executing a trained model requires significant computational resources incurring monetary and environmental costs. We propose a modification to contextual representation fine-tuning which, during inference, allows for an early (and fast) "exit" We test our proposed modification on five different datasets in two tasks: three text classification datasets and two natural language inference benchmarks.
  arXiv  Detail & Related papers  (2020-04-16T04:28:08Z)

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.