GraphHCA: Closed-Form Hindsight Credit Assignment for Long-Horizon LLM Agents

arXiv:2609.35084 · cs.LG · Submitted 2026-09-28 · Read on arXiv

cs.LG

Submitted: 2026-09-28

Updated: 2026-09-28

Code: https://github.com/mpSchrader/gym-sokoban

Terminology

• Credit assignment — 1× in this paper · explained in 3 episodes

Sources

• GPT-4 Technical Report
• Qwen2.5-VL Technical Report
• Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
• Beyond Trajectory-Level Attribution: Graph-Based Credit Assignment for Agentic Reinforcement Learning
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
• Kimi k1.5: Scaling Reinforcement Learning with LLMs
• Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning
• Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents
• Proximal Policy Optimization Algorithms
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Hindsight Credit Assignment for Long-Horizon LLM Agents
• Gemini: A Family of Highly Capable Multimodal Models
• Qwen2.5 Technical Report
• Fine-Tuning Language Models from Human Preferences

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks