Self-Evolving Deep Research via Joint Generation and Evaluation

• URL: http://arxiv.org/abs/2606.04507v1
• Date: Wed, 03 Jun 2026 06:38:19 GMT
• Title: Self-Evolving Deep Research via Joint Generation and Evaluation
• Abstract summary: Deep research report generation lacks definitive ground-truth, making reward design inherently unverifiable.<n>Existing approaches mitigate this challenge with LLM-as-a-judge and query-dependent evaluation rubrics.<n>We address this limitation with a textbfself-evolving textbfco-evolutionary training framework for deep textbfresearch evaluation and generation.
• Score: 29.861012849163945
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Large Language Models (LLMs) have become increasingly adopted in daily applications, with deep research standing out as a particularly important capability. Unlike traditional question-answering (QA) tasks, deep research report generation lacks definitive ground-truth, making reward design inherently unverifiable and limiting effective reinforcement learning. Existing approaches mitigate this challenge with LLM-as-a-judge and query-dependent evaluation rubrics, but they still rely on static evaluators that cannot adapt their standards as the solver improves, leading to insufficient and eventually saturated optimization pressure. We address this limitation with a \textbf{s}elf-evolving \textbf{co}-evolutionary training framework for deep \textbf{re}search evaluation and generation (SCORE), which tightly couples an evaluator and a solver in a shared-parameter learning process. Rather than treating generation and evaluation as isolated modules, we leverage their intrinsic connection to enable joint improvement within a single shared-parameter model. To restrict this process, we introduce a meta-harness, which dynamically controls the evaluation environment based on solver performance, encouraging valid evaluation dimensions and sufficiently deep evaluator search. Extensive experiments on deep research benchmarks demonstrate consistent improvement in report generation quality, showing that co-evolving evaluation and generation is a promising direction for training open-ended research agents.

Related papers

• LigBench: A Unified and Human-Aligned Benchmark for LLM-based Research Idea Generation [36.122822391024755]
  LigBench is an automated evaluation benchmark that enables fine-grained and reliable evaluation of AI research ideas.<n>P PAIR-IQ is a dataset tailored for training pairwise idea judgment models.<n>Experiments demonstrate that LigBench achieves stable and interpretable evaluations.
  arXiv  Detail & Related papers  (2026-08-13T12:11:23Z)
• LLM-Based Scientific Peer Review: Methods, Benchmarks, and Reliability Challenges [4.655159257282136]
  This survey offers a systems-level analysis of large language models (LLMs)-based scientific peer review.<n>We focus on two core evaluative functions: critique generation and score prediction.<n>We identify emerging risks, including prompt injection, data poisoning, retrieval vulnerabilities, and reward hacking.
  arXiv  Detail & Related papers  (2026-06-23T18:12:39Z)
• From Holistic Evaluation to Structured Criteria: Rubrics Across the Evolving LLM Landscape [79.30826980815927]
  rubrics are explicit criteria sets that transform complex quality judgments into structured and actionable standards.<n>We systematically organize existing rubric designs, examine their construction and optimization, and analyze their role across evaluation and training.
  arXiv  Detail & Related papers  (2026-06-07T13:34:55Z)
• Search-R2: Enhancing Search-Integrated Reasoning via Actor-Refiner Collaboration [49.9937230730202]
  We propose Search-R2, a novel Actor-Refiner collaboration framework that enhances reasoning through targeted intervention.<n>Our approach decomposes the generation process into an Actor, which produces initial reasoning trajectories.<n>We show that Search-R2 consistently outperforms strong RAG and RL-based baselines across model scales.
  arXiv  Detail & Related papers  (2026-02-03T15:32:09Z)
• SelfAI: Building a Self-Training AI System with LLM Agents [79.10991818561907]
  SelfAI is a general multi-agent platform that combines a User Agent for translating high-level research objectives into standardized experimental configurations.<n>An Experiment Manager orchestrates parallel, fault-tolerant training across heterogeneous hardware while maintaining a structured knowledge base for continuous feedback.<n>Across regression, computer vision, scientific computing, medical imaging, and drug discovery benchmarks, SelfAI consistently achieves strong performance and reduces redundant trials.
  arXiv  Detail & Related papers  (2025-11-29T09:18:39Z)
• JudgeAgent: Knowledge-wise and Dynamic LLM Evaluation with Agent-as-Interviewer [19.09571232466437]
  We propose Agent-as-Interviewer, a dynamic evaluation paradigm for large language models (LLMs)<n>Unlike current benchmarking or dynamic interaction paradigms, Agent-as-Interviewer utilizes agents to invoke knowledge tools for wider and deeper knowledge in the dynamic multi-turn question generation.<n>We develop JudgeAgent, a knowledge-wise dynamic evaluation framework that employs knowledge-driven synthesis as the agent's tool and uses difficulty scoring as strategy guidance.
  arXiv  Detail & Related papers  (2025-09-02T08:52:16Z)
• Retrieval-Augmented Generation: A Comprehensive Survey of Architectures, Enhancements, and Robustness Frontiers [0.0]
  Retrieval-Augmented Generation (RAG) has emerged as a powerful paradigm to enhance large language models.<n>RAG introduces new challenges in retrieval quality, grounding fidelity, pipeline efficiency, and robustness against noisy or adversarial inputs.<n>This survey aims to consolidate current knowledge in RAG research and serve as a foundation for the next generation of retrieval-augmented language modeling systems.
  arXiv  Detail & Related papers  (2025-05-28T22:57:04Z)
• RAG-Zeval: Towards Robust and Interpretable Evaluation on RAG Responses through End-to-End Rule-Guided Reasoning [64.46921169261852]
  RAG-Zeval is a novel end-to-end framework that formulates faithfulness and correctness evaluation as a rule-guided reasoning task.<n>Our approach trains evaluators with reinforcement learning, facilitating compact models to generate comprehensive and sound assessments.<n>Experiments demonstrate RAG-Zeval's superior performance, achieving the strongest correlation with human judgments.
  arXiv  Detail & Related papers  (2025-05-28T14:55:33Z)
• ParamMute: Suppressing Knowledge-Critical FFNs for Faithful Retrieval-Augmented Generation [91.20492150248106]
  We investigate the internal mechanisms behind unfaithful generation and identify a subset of mid-to-deep feed-forward networks (FFNs) that are disproportionately activated in such cases.<n>We propose Parametric Knowledge Muting through FFN Suppression (ParamMute), a framework that improves contextual faithfulness by suppressing the activation of unfaithfulness-associated FFNs.<n> Experimental results show that ParamMute significantly enhances faithfulness across both CoFaithfulQA and the established ConFiQA benchmark, achieving substantial reductions in reliance on parametric memory.
  arXiv  Detail & Related papers  (2025-02-21T15:50:41Z)
• DeepCRCEval: Revisiting the Evaluation of Code Review Comment Generation [11.010557279355885]
  This study empirically analyzes benchmark comments using a novel set of criteria informed by prior research and developer interviews.<n>Our evaluation framework, DeepCRCEval, integrates human evaluators and Large Language Models (LLMs) for a comprehensive reassessment of current techniques.
  arXiv  Detail & Related papers  (2024-12-24T08:53:54Z)
• Progress or Regress? Self-Improvement Reversal in Post-training [26.051637877066327]
  We propose a comprehensive evaluative framework to scrutinize the underlying enhancements of post-training paradigms for self-improvement. We show that models showing improved performance across benchmarks will paradoxically exhibit declines in broader, essential capabilities. These findings indicate that current self-improvement practices through post-training are inadequate for equipping models to tackle more complex problems.
  arXiv  Detail & Related papers  (2024-07-06T09:07:11Z)
• Position: AI Evaluation Should Learn from How We Test Humans [65.36614996495983]
  We argue that psychometrics, a theory originating in the 20th century for human assessment, could be a powerful solution to the challenges in today's AI evaluations.
  arXiv  Detail & Related papers  (2023-06-18T09:54:33Z)
• Towards Automatic Evaluation of Dialog Systems: A Model-Free Off-Policy Evaluation Approach [84.02388020258141]
  We propose a new framework named ENIGMA for estimating human evaluation scores based on off-policy evaluation in reinforcement learning. ENIGMA only requires a handful of pre-collected experience data, and therefore does not involve human interaction with the target policy during the evaluation. Our experiments show that ENIGMA significantly outperforms existing methods in terms of correlation with human evaluation scores.
  arXiv  Detail & Related papers  (2021-02-20T03:29:20Z)

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.