Awesome Auto Research Hub Papers · Datasets · Projects
← Back to papers

ARISE: Agentic Rubric-Guided Iterative Survey Engine for Automated Scholarly Paper Generation

arXiv 2025 73.2 method

TLDR

ARISE uses specialized LLM agents and rubric-guided iterative refinement to automatically generate high-quality academic survey papers.

Reasoning

The paper presents a novel modular architecture with role-specific agents and a structured feedback loop, achieving strong quantitative results. However, it focuses narrowly on survey generation rather than broader scientific discovery, and the abstract lacks details on limitations or generalizability.

Read-first score

Read-first score 73.2, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 67.

Methodology quality 25%
90

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=baseline,evaluation,experiment,metric,result

Recency 8%
86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Reproducibility 25%
81

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=True; dataset=False; markers=code,github

Topical relevance 42%
55.8

Uses existing LLM keyword relevance scores normalized to 0-100. AI scientist,automated scientific discovery,autonomous research agent,automated research,literature review agent,survey generation,automated experimentation,experiment design agent,AI for scientific research,paper writing agent,research automation,scientific discovery agent

Field roles

FrontierMethodology anchorReproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 47.

Keyword Scores

survey generation
10
literature review agent
9
automated research
8
paper writing agent
8
autonomous research agent
7
research automation
7
AI for scientific research
6
automated scientific discovery
4
AI scientist
3
scientific discovery agent
3
automated experimentation
1
experiment design agent
1

Deep Analysis

Innovations

  • Agentic rubric-guided iterative refinement loop with multiple reviewer agents
  • Modular architecture of specialized LLM agents mirroring distinct scholarly roles (topic expansion, citation curation, summarization, drafting, peer review)
  • Behaviorally anchored rubric for structured multi-agent evaluation

Methodology

ARISE employs a modular agentic system where LLM agents perform distinct scholarly tasks, and a rubric-guided iterative refinement loop with multiple reviewer agents that assess drafts using a structured, behaviorally anchored rubric, synthesizing feedback to improve the manuscript. Evaluation compares ARISE against state-of-the-art automated systems and recent human-written surveys.

Key Results

ARISE achieved an average rubric-aligned quality score of 92.48, consistently outperforming baseline methods in comprehensiveness, accuracy, formatting, and scholarly rigor.

Tags

DLAI