inference-net / Aella-Qwen3-14B

Paper-Summarizer-Qwen3-14B

A fine-tuned Qwen3-14B model specialized for generating structured summaries of scientific research papers in standardized JSON format.

Model Description

This model is part of Project AELLA , developed in collaboration with LAION and Wynd Labs to democratize access to scientific knowledge by creating structured summaries of research papers at scale.

Base Model : Qwen 3 14B Training Data : 110,000 curated research papers Performance : Achieves 73.9% accuracy on QA evaluation, comparable to GPT-5 (74.6%) Cost Efficiency : 98% lower cost than closed-source alternatives

This generates comprehensive structured summaries in a JSON format. The papers are either classified as SCIENTIFIC_TEXT, PARTIAL_SCIENTIFIC_TEXT, or NON_SCIENTIFIC_TEXT. The fields extracted are key research elements such as methodology, results, claims, and limitations.

The model supports papers up to 131K tokens.

Usage

Serving the Model

vllm serve inference-net/Paper-Summarizer-Qwen3-14B \
--port 8000 \
--host 0.0.0.0 \
--trust-remote-code \
--data-parallel-size 1 \
--tensor-parallel-size 1 \
--max-num-seqs 32 \
--max-model-len 131072 \
--max-num-batched-tokens 8192 \
--gpu-memory-utilization 0.90 \
--enable-prefix-caching \
--enable-chunked-prefill

Making Requests

import requests

# System prompt (required)
system_prompt = """[Insert the full system prompt from the prompt.txt file -
see the full prompt in the model repository]"""

# User prompt: the paper text to summarize
paper_text = """
Title: Your Paper Title
Authors: Author 1, Author 2
Abstract: ...
[Full paper content]
"""

# API request
response = requests.post(
"http://localhost:8000/v1/chat/completions",
json={
"model": "inference-net/Paper-Summarizer-Qwen3-14B",
"messages": [
{"role": "system", "content": system_prompt},
{"role": "user", "content": paper_text}
],
"temperature": 0.2
},
timeout=600
)

result = response.json()
summary = result["choices"][0]["message"]["content"]
print(summary)

System Prompt

The model requires a specific system prompt that defines the JSON schema and extraction instructions. The prompt instructs the model to:

1. Classify the text as SCIENTIFIC_TEXT, PARTIAL_SCIENTIFIC_TEXT, or NON_SCIENTIFIC_TEXT
2. Extract structured information including:

• Title, authors, publication year
• Research context and hypotheses
• Methodological details
• Key results with quantitative data
• Claims with supporting evidence
• Limitations and ethical considerations

The full system prompt is available in the model repository's prompt.txt file.

Output Format

The model outputs a single valid JSON object with this structure:

{
"article_classification": "SCIENTIFIC_TEXT",
"reason": null,
"summary": {
"title": "",
"authors": "",
"publication_year": null,
"field_subfield": "",
"executive_summary": "",
"research_context": "",
"methodological_details": "",
"key_results": "",
"claims": [...],
"contradictions_and_limitations": "",
...
}
}

Performance

LLM-as-a-Judge Evaluation

• Score : 4.207/5.0
• Comparison : Within 15% of GPT-5 (4.805/5.0)

QA Dataset Evaluation

• Accuracy : 73.9%
• Comparison : Ties with Gemini 2.5 Flash, nearly matches GPT-5 (74.6%)

Throughput

• Requests/sec : 0.43
• Input Tokens/sec : 7,516.54
• Output Tokens/sec : 2,588.30

Training Details

• Training Set : 100,000 papers
• Validation Set : 10,000 papers
• Average Paper Length : 81,334 characters
• Training Approach : Post-training on summaries generated by frontier models (GPT-5, Claude 4.5 Sonnet, Gemini 2.5 Pro)

Limitations

• May generate subtle factual errors (hallucinations) for fine-grained details
• Context limit (131K tokens) may truncate extremely long documents
• Unified schema may not capture all domain-specific nuances
• Summaries are research aids, not replacements for primary sources in high-stakes scenarios

Related Resources

• Paper Visualization Website : https://laion.inference.net
• Visualization Repository : https://github.com/context-labs/laion-data-explorer
• Alexandria Paper : https://arxiv.org/abs/2502.19413
• Nemotron Variant : inference-net/Paper-Summarizer-Nemotron-12B

License

[License information to be added]

Acknowledgments

This work was made possible through collaboration with:

• LAION
• Wynd Labs
• Inference.net
• Contributors to bethgelab, PeS2o, Common Pile, and OpenAlex