paper-banana（纸香蕉）

PaperBanana: Academic Illustration Pipeline

Automates publication-ready academic illustrations via 5 specialized agents, each implemented as a separate Gemini API call: Retriever (categorize & select references) -> Planner (multimodal description) -> Stylist (polish) -> Visualizer (render) -> Critic (evaluate & refine).

Two output modes:

• DIAGRAM MODE: Each agent is a Python script calling Gemini VLM/image APIs. Run scripts/orchestrate.py for end-to-end execution.
• PLOT MODE: Statistical plots generated as executable Python matplotlib/seaborn code (code-based to eliminate data hallucination).

Requirements: GOOGLE_API_KEY env var (used for VLM calls in retriever/planner/stylist/critic AND image generation in visualizer), Python 3.10+ with google-genai, matplotlib, seaborn, numpy, pillow.

Paper: *PaperBanana: Automating Academic Illustrations with Multi-Agent Systems* (arXiv:2601.23265, Google/PKU)

Step 1: Determine Output Mode

Decide which track to follow:

Critical rule: PLOT MODE always generates Python code (never image generation for data visualizations). Code-based generation eliminates data hallucination errors that corrupt numerical accuracy in image-based approaches.

Step 2: Execute Pipeline

DIAGRAM MODE — Automated Pipeline

Primary entry point: Run the end-to-end orchestrator:

python scripts/orchestrate.py \
  --methodology-file methodology.txt \
  --caption "Figure 1: Overview of proposed framework" \
  --mode diagram \
  --output output/diagram.png

Or with inline text:

python scripts/orchestrate.py \
  --methodology "Our framework consists of three modules..." \
  --caption "Figure 1: System overview" \
  --mode diagram \
  --output output/diagram.png

The orchestrator chains all 5 agents automatically and handles the Critic's refinement loop (up to 3 iterations). Intermediate outputs are saved to output/work/ for inspection.

Pipeline Details

Read references/DIAGRAM-PROMPTS.md for the actual Gemini prompt templates used by each agent.

Phase 1: RETRIEVER (scripts/retriever.py) — Gemini VLM call

• Classifies methodology into 1 of 4 categories from references/DIAGRAM-CATEGORIES.md
• Selects 2 most relevant reference diagrams from the 13 curated examples in assets/references/
• Identifies visual intent: Framework Overview, Pipeline/Flow, Detailed Module, Architecture Diagram

Phase 2: PLANNER (scripts/planner.py) — Multimodal Gemini VLM call

• Sends the 2 selected reference images + methodology text to Gemini as a multimodal prompt
• The VLM "sees" what good methodology diagrams look like (in-context learning from images)
• Generates an extremely detailed textual description of the target diagram
• Critical: Natural language only for all visual attributes. NEVER hex codes or pixel dimensions

Phase 3: STYLIST (scripts/stylist.py) — Gemini VLM call

• Takes the Planner's description + full NeurIPS 2025 style guide
• Applies domain-specific styling based on the category from Phase 1
• Follows 5 critical rules: preserve aesthetics, intervene minimally, respect domain, enrich details, preserve content
• Outputs the polished description only

Phase 4: VISUALIZER (scripts/generate_image.py) — Gemini Image API call

• Uses gemini-3-pro-image-preview to generate the diagram image from the styled description
• Prepends quality prefix (high-res, legible text, clean background, no watermarks)
• Aspect ratio selected based on visual intent (16:9 for pipelines, 3:2 for modules)

Phase 5: CRITIC (scripts/critic.py) — Multimodal Gemini VLM call

• Sends the generated image + methodology text to Gemini for multimodal evaluation
• Scores on 4 dimensions (faithfulness, readability, conciseness, aesthetics)
• If faithfulness < 7 OR readability < 7: generates revised description → loops to Phase 4
• Maximum 3 refinement iterations

DIAGRAM MODE — Manual Execution

You can also run each agent individually for more control:

# Phase 1: Retriever
python scripts/retriever.py --methodology-file text.txt --output work/retriever.json

# Phase 2: Planner
python scripts/planner.py --methodology-file text.txt --caption "Figure 1: ..." \
  --references work/retriever.json --output work/planner.json

# Phase 3: Stylist
python scripts/stylist.py --description work/planner.json --output work/stylist.json

# Phase 4: Visualizer (extract styled_description from JSON, pass to generate_image.py)
python scripts/generate_image.py --prompt-file work/styled_desc.txt --output output/diagram.png

# Phase 5: Critic
python scripts/critic.py --image output/diagram.png --methodology-file text.txt \
  --description work/stylist.json --output work/critic.json

PLOT MODE

Read references/PLOT-PROMPTS.md for detailed agent prompts. Read references/PLOT-STYLE-GUIDE.md for aesthetic rules.

Plot mode uses Claude (or the host agent) for reasoning and code generation — no Gemini API calls needed for plot generation itself.

Phase 1: CATEGORIZE (Retriever)

Match data characteristics and visual intent:

Phase 2: PLAN (Planner)

Create a detailed specification that explicitly enumerates:

• Every raw data point with exact coordinates/values
• Axis ranges, labels, tick marks, scales (linear/log)
• Color assignments for each series/category
• Font sizes for title, axis labels, tick labels, legend
• Line widths, marker sizes, marker shapes
• Legend placement and formatting
• Grid style (major/minor, dashed/solid)
• Figure dimensions and DPI

Phase 3: STYLE (Stylist)

Read references/PLOT-STYLE-GUIDE.md for NeurIPS 2025 plot aesthetics.

Key styling rules:

• White backgrounds only
• Colorblind-friendly palettes (see assets/palettes/colorblind_safe.json)
• Sans-serif fonts (Helvetica, Arial, or DejaVu Sans)
• Markers on line charts for print readability
• Inward-facing tick marks
• Subtle grid lines (light gray, dashed)

Phase 4: VISUALIZE (Visualizer — Code Generation)

Generate complete, self-contained Python matplotlib/seaborn code. Use scripts/plot_generator.py as a reference implementation or run it directly with a JSON config:

python scripts/plot_generator.py --config plot_config.json --output figure.pdf

Code requirements:

• Self-contained: all data defined inline, no external file dependencies
• Apply .mplstyle from assets/matplotlib_styles/academic_default.mplstyle
• Set OUTPUT_PATH variable for output file location
• 300 DPI, bbox_inches='tight'
• No plt.show() — save only
• Support both PDF and PNG output

After generating the code, execute it to produce the plot image.

Phase 5: CRITIQUE (Critic)

Same rubric as diagram mode, plus plot-specific checks:

• Data fidelity: Every data point correctly plotted
• Axis accuracy: Ranges, labels, scales match specification
• Layout: No overlapping labels, legends, or data points
• Code correctness: Syntax valid, imports available, output saved

If code execution failed, analyze the error, simplify the approach, and regenerate.

Quick Start Examples

Diagram (automated): Run scripts/orchestrate.py with your methodology text file and caption.

Diagram (via agent): "Generate a methodology diagram for my transformer architecture. Here is the methodology section: [paste text]. Caption: Overview of our proposed multi-head attention framework."

Plot: "Create a bar chart comparing model performance. Data: {BERT: 92.3, GPT-4: 88.1, Claude: 95.7, Gemini: 91.2}. Intent: F1 score comparison across language models."

Improve: "Improve the aesthetics of this diagram: [paste existing description or attach current figure]"

File Reference

Environment Setup

# Required for all Gemini API calls (VLM reasoning + image generation)
export GOOGLE_API_KEY="your-api-key-here"

# Install dependencies
pip install google-genai matplotlib seaborn numpy pillow

Verify setup: python scripts/validate_output.py --check-deps