QuantEvolver applies reinforcement fine-tuning to evolve an LLM policy for generating executable alpha factor expressions, yielding higher-quality and more complementary factors than prompt-based baselines on market benchmarks.
d-TreeRPO: Towards More Reliable Policy Optimization for Diffusion Language Models
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
abstract
Reinforcement learning (RL) is pivotal for enhancing the reasoning capabilities of diffusion large language models (dLLMs). However, existing dLLM policy optimization methods suffer from two critical reliability bottlenecks: (1) reward sparsity, arising from coarse or unverifiable signals that impede accurate advantage calculation; and (2) their probability estimates do not account for the gap to the unbiased expectation over all decoding orders, which are intractable to compute. To mitigate these issues, we propose d-TreeRPO, a reliable RL framework for dLLMs that leverages tree-structured rollouts and bottom-up advantage computation based on verifiable outcome rewards to provide fine-grained and verifiable step-wise reward signals. Furthermore, we provide a theoretical proof demonstrating that increasing prediction confidence effectively minimizes the gap between unbiased expected prediction probabilities and its single-step forward pass estimate. Guided by this analysis, we introduce a time-scheduled self-distillation loss during training that enhances prediction confidence in later training stages, thereby enabling more accurate probability estimation and better performance. Experiments demonstrate that d-TreeRPO outperforms existing baselines and achieves significant improvements across multiple reasoning benchmarks. Specifically, it achieves +86.2% on Sudoku, +51.6% on Countdown, +4.5% on GSM8K, and +5.3% on Math500 compared to the base model.
citation-role summary
background 3
citation-polarity summary
years
2026 7roles
background 3polarities
background 3representative citing papers
RSPO interprets reward advantages as targets for relative log-ratios in dLLMs, calibrating noisy estimates to stabilize RLVR training and achieve strong gains on planning tasks with competitive math reasoning performance.
Introduces the first benchmark for fine-grained failures in reinforcement fine-tuning of LLMs and an automatic management framework that detects, diagnoses, and remediates them.
E2E-REME outperforms nine LLMs in accuracy and efficiency for end-to-end microservice remediation by using experience-simulation reinforcement fine-tuning on a new benchmark called MicroRemed.
DMax uses On-Policy Uniform Training and Soft Parallel Decoding to enable aggressive parallelism in dLLMs, raising TPF on GSM8K from 2.04 to 5.47 and on MBPP from 2.71 to 5.86 while preserving accuracy.
Proposes HT-GRPO with sketch-then-paint staged updates, prompt-conditioned importance ratios, and hierarchical credit assignment for dMLLMs, reporting gains on GenEval and DPG plus quality metrics.
RCLAgent uses multi-agent recursion-of-thought with parallel reasoning on trace graphs to outperform prior methods in root cause localization accuracy and efficiency for microservice systems.
citing papers explorer
-
From Feedback Loops to Policy Updates: Reinforcement Fine-Tuning for LLM-Based Alpha Factor Discovery
QuantEvolver applies reinforcement fine-tuning to evolve an LLM policy for generating executable alpha factor expressions, yielding higher-quality and more complementary factors than prompt-based baselines on market benchmarks.
-
Relative Score Policy Optimization for Diffusion Language Models
RSPO interprets reward advantages as targets for relative log-ratios in dLLMs, calibrating noisy estimates to stabilize RLVR training and achieve strong gains on planning tasks with competitive math reasoning performance.
-
Towards Robust LLM Post-Training: Automatic Failure Management for Reinforcement Fine-Tuning
Introduces the first benchmark for fine-grained failures in reinforcement fine-tuning of LLMs and an automatic management framework that detects, diagnoses, and remediates them.
-
E2E-REME: Towards End-to-End Microservices Auto-Remediation via Experience-Simulation Reinforcement Fine-Tuning
E2E-REME outperforms nine LLMs in accuracy and efficiency for end-to-end microservice remediation by using experience-simulation reinforcement fine-tuning on a new benchmark called MicroRemed.
-
DMax: Aggressive Parallel Decoding for dLLMs
DMax uses On-Policy Uniform Training and Soft Parallel Decoding to enable aggressive parallelism in dLLMs, raising TPF on GSM8K from 2.04 to 5.47 and on MBPP from 2.71 to 5.86 while preserving accuracy.
-
Sketch Then Paint: Hierarchical Reinforcement Learning for Diffusion Multi-Modal Large Language Models
Proposes HT-GRPO with sketch-then-paint staged updates, prompt-conditioned importance ratios, and hierarchical credit assignment for dMLLMs, reporting gains on GenEval and DPG plus quality metrics.
-
Towards In-Depth Root Cause Localization for Microservices with Multi-Agent Recursion-of-Thought
RCLAgent uses multi-agent recursion-of-thought with parallel reasoning on trace graphs to outperform prior methods in root cause localization accuracy and efficiency for microservice systems.