verifiable-rewards

This paper introduces SARA, a sequential adaptive rollout allocation method for RLVR that abandons saturated groups early and reallocates the budget, achieving comparable accuracy with 22% fewer rollouts than dynamic sampling and up to 67% savings when combined.

This paper proposes RLSVR, a task-transformation paradigm that extends reinforcement learning with verifiable rewards to open-ended LLM tasks by creating self-verifiable proxy environments, instantiated via the SpyRL multi-agent self-play framework, showing gains on summarization, creative writing, and math reasoning.

This paper diagnoses 'pass@k inversion' in reinforcement learning with verifiable rewards (RLVR), where training improves one-shot accuracy but reduces performance under repeated sampling, particularly on boundary prompts where correct trajectories are rare. It proposes a per-problem base anchoring (PBA) method to mitigate this issue.

Introduces Surrogate Latent Policy Optimization (SLPO) to apply outcome-reward RL to autoregressive latent reasoners, enabling test-time scaling and variable-horizon policies that improve accuracy on harder instances.

This paper introduces a framework using reinforcement learning from verifiable rewards to train large language models for strategic bargaining in multi-buyer markets, addressing private information and surplus extraction in concurrent negotiations.

This paper systematically compares process and outcome reward structures for reinforcement learning with verifiable rewards (RLVR) in small language models for mathematical reasoning. The study finds that process-only supervision significantly improves accuracy and reasoning trace fidelity over outcome-only supervision, and analyzes failure modes.

This paper presents a proof-of-concept using Reinforcement Learning with Verifiable Rewards (RLVR) to train small language models for tool-use in enterprise SaaS workflows like Jira and Confluence. The approach uses synthetic environments and GRPO training to improve tool-call accuracy, achieving significant reward gains over baselines.

The paper proposes a verifiable label-free reward for training calibrated probabilistic forecasters using reinforcement learning, avoiding the calibration degradation that occurs when rewarding single outcomes. Applied to NFL win probability, a 7B model trained with this reward achieves calibration comparable to the betting market.

BV-Blend is a critic-free reinforcement learning framework that combines prompt-local on-policy statistics with historical moments from semantic clusters to stabilize advantage estimation, improving training stability and performance for aligning large language models with verifiable rewards.

Proposes Tandem Reinforcement Learning (TRL), extending the tandem training paradigm to RLVR to improve reasoning compatibility and legibility for weaker models and humans, showing that TRL matches solo performance while enhancing handoff robustness and reducing distributional drift.

Reasoning Arena improves reinforcement learning with verifiable rewards by using trace tournaments and Bradley-Terry models to generate meaningful gradients from non-diverse reward groups, resulting in faster training and better reasoning performance.

Proposes CVT-RL, a constrained policy-gradient algorithm with policy-conditioned counterfactual contribution estimation and verifiable rewards, improving long-horizon language agent reliability and reducing reward hacking.

Introduces Atomic Decomposition and Recombination (ADR), a framework that generates novel and challenging verifiable code tasks by decomposing and recombining atomic elements, enabling scalable reinforcement learning with verifiable rewards for large language models.

CUA-Gym introduces a scalable pipeline for generating verifiable training environments and tasks for computer-use agents, addressing data scarcity. The resulting dataset and models achieve strong performance on benchmarks like OSWorld-Verified and WebArena.

Introduces DelTA, a discriminative token credit assignment method for reinforcement learning from verifiable rewards (RLVR) that amplifies distinctive token-gradient directions and reduces noise from shared patterns, achieving significant improvements on mathematical and code generation benchmarks.

An analysis of why RL for coding tasks is gaining traction due to verifiable rewards, and why the emerging framework Harbor addresses the bottleneck of environment complexity in RL training.

VideoRLVR optimizes video diffusion models for verifiable reasoning tasks using reinforcement learning with rule-based rewards, achieving better performance than supervised methods in constraint-satisfying video generation.

This paper introduces Listwise Policy Optimization (LPO), a method for RLVR that explicitly handles target projection via divergence minimization on the response simplex to improve training stability and performance in LLMs.

Researchers from Tianjin University and Alibaba Group propose EA-RLVR, a reinforcement learning framework with verifiable rewards that improves cross-cultural entity translation in LLMs by activating parametric knowledge already encoded during pre-training, without relying on external knowledge bases. Training on 7k samples boosts Qwen3-14B's entity translation accuracy from 23.66% to 31.87% on unseen entities.

Huggingface introduces EcomRLVE-GYM, a framework providing eight verifiable environments for training reinforcement learning agents on complex e-commerce tasks. The tool features adaptive difficulty curricula and algorithmic rewards to improve task completion in shopping assistants, demonstrated by training a Qwen 3 8B model.