# STAMP: Provenance-Guided Credit Assignment for Deep Search Agents

arXiv: https://arxiv.org/abs/2607.11172 · HTML: https://mlsanity.com/paper/2607.11172/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 31 references that are in the map.
- First posted on arXiv 13 Jul 2026.

## Builds on (most cited first, 18 of 18)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [Hindsight Experience Replay](https://mlsanity.com/paper/1707.01495/) (2017)
- [BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agents](https://mlsanity.com/paper/2504.12516/) (2025)
- [The Lessons of Developing Process Reward Models in Mathematical Reasoning](https://mlsanity.com/paper/2501.07301/) (2025)
- [WebSailor: Navigating Super-human Reasoning for Web Agent](https://mlsanity.com/paper/2507.02592/) (2025)
- [Beyond Ten Turns: Unlocking Long-Horizon Agentic Search with Large-Scale Asynchronous RL](https://mlsanity.com/paper/2508.07976/) (2025)
- [BrowseComp-ZH: Benchmarking Web Browsing Ability of Large Language Models in Chinese](https://mlsanity.com/paper/2504.19314/) (2025)
- [WebShaper: Agentically Data Synthesizing via Information-Seeking Formalization](https://mlsanity.com/paper/2507.15061/) (2025)
- [DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning](https://mlsanity.com/paper/2511.22570/) (2025)
- [WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents](https://mlsanity.com/paper/2509.06501/) (2025)
- [Process Reward Models for LLM Agents: Practical Framework and Directions](https://mlsanity.com/paper/2502.10325/) (2025)
- [Process vs. Outcome Reward: Which is Better for Agentic RAG Reinforcement Learning](https://mlsanity.com/paper/2505.14069/) (2025)
- [DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RL](https://mlsanity.com/paper/2509.10446/) (2025)
- [Dynamic and Generalizable Process Reward Modeling](https://mlsanity.com/paper/2507.17849/) (2025)
- [Repurposing Synthetic Data for Fine-grained Search Agent Supervision](https://mlsanity.com/paper/2510.24694/) (2025)

## Built on by (most cited first, 5 of 5)

- [HERALD: Counterfactual Audits and Minimal Repairs for Proof-of-Retrieval Rewards](https://mlsanity.com/paper/2608.06012/) (2026)
- [Beyond Outcome Rewards: Constructing and Assigning Retrieval Credit for Search Agents](https://mlsanity.com/paper/2610.10179/) (2026)
- [OneSearch-VL: Unified Multimodal Deep Research Agent for Image and Video](https://mlsanity.com/paper/2610.12419/) (2026)
- [PANINI-DR: Controlled Deep Research through Generative Semantic Workspaces](https://mlsanity.com/paper/2610.15208/) (2026)
- [WorkRover: Scaling Long-Horizon Cowork Agent Training Via Recursive Environment Evolution](https://mlsanity.com/paper/2610.18067/) (2026)

## Abstract (arXiv)

Reinforcement learning for deep-search agents has largely focused on trajectory-level scoring -- outcome correctness, citation-aware rewards, and evidence coverage. Yet the actions that expose supporting documents receive no targeted credit, a gap we call the reward-credit mismatch. We propose STAMP, in which a reference-based verifier judges whether each cited document supports an entity or relation in a training-time evidence graph, and first-exposure attribution traces each supported citation back to the action that first surfaced it. This step credit is injected through sign-preserving advantage modulation, which redistributes advantage across steps without changing the trajectory-level reward or the relative ranking of trajectories within each group. On BrowseComp, BrowseComp-ZH, and xbench-DS, STAMP improves the GRPO baseline by +2.0/+5.5/+3.0 points under matched SFT initialization, training data, and search tools, and composes with both outcome-only and citation-rubric base rewards. Component ablations confirm that the provenance-based credit signal and the sign-preserving advantage modulation each contribute to the gains.