API Documentation

{
  "status": "ok",
  "plan": "free",
  "key_name": "my-research-key",
  "usage_today": 47
}

{
  "papers": [
    {
      "arxiv_id": "2401.04088",
      "title": "Attention Is All You Need (But Not All You Get)",
      "authors": ["A. Researcher", "B. Scientist"],
      "year": 2024,
      "categories": ["cs.LG", "cs.AI"],
      "primary_category": "cs.LG",
      "arxiv_url": "https://arxiv.org/abs/2401.04088",
      "has_code": true,
      "github_url": "https://github.com/example/repo",
      "citation_count": 42,
      "rank_score": 0.73,
      "llm_summary": "Proposes a sparse attention variant that reduces compute by 60% while matching dense attention accuracy on 5 benchmarks.",
      "llm_novelty_score": 0.55
    }
  ],
  "total": 1847,
  "page": 1,
  "limit": 20,
  "next_cursor": "eyJzIjogMC43MywgImlkIjogIjI0MDEuMDQwODgifQ=="
}

{
  "papers": [
    { "arxiv_id": "2401.04088", "title": "Attention Is All You Need (But Not All You Get)", "rank_score": 0.73 },
    { "arxiv_id": "2312.08901", "title": "Efficient Sparse Transformers for Long Sequences", "rank_score": 0.68 }
  ],
  "total": 2,
  "not_found": ["9999.99999"]
}

{
  "arxiv_id": "2401.04088",
  "title": "Attention Is All You Need (But Not All You Get)",
  "authors": ["A. Researcher", "B. Scientist"],
  "year": 2024,
  "categories": ["cs.LG", "cs.AI"],
  "primary_category": "cs.LG",
  "arxiv_url": "https://arxiv.org/abs/2401.04088",
  "pdf_url": "https://arxiv.org/pdf/2401.04088",
  "has_code": true,
  "github_url": "https://github.com/example/repo",
  "citation_count": 42,
  "rank_score": 0.73,
  "llm_summary": "Proposes a sparse attention variant that reduces compute by 60% while matching dense attention accuracy on 5 benchmarks.",
  "llm_significance": "Significant for large-scale transformer training; reduces memory cost without accuracy loss.",
  "llm_novelty_score": 0.55,
  "venue_name": "NeurIPS"
}

{
  "papers": [
    {
      "arxiv_id": "2312.08901",
      "title": "Efficient Sparse Transformers for Long Sequences",
      "similarity_score": 0.87
    }
  ],
  "total": 10
}

{
  "papers": [
    {
      "arxiv_id": "2403.11234",
      "title": "Scaling Laws for Sparse Attention"
    }
  ],
  "total": 15,
  "direction": "cited_by"
}

{
  "papers": [
    {
      "arxiv_id": "2406.11409",
      "title": "CodeGemma: Open Code Models Based on Gemma",
      "llm_summary": "...",
      "similarity_score": 0.702
    },
    {
      "arxiv_id": "2406.11925",
      "title": "DocCGen: Document-based Controlled Code Generation",
      "similarity_score": 0.7009
    }
  ],
  "total": 2,
  "direction": "cited_by",
  "query": "code generation"
}

{
  "arxiv_id": "2401.04088",
  "results_text": "Our method achieves 94.2% accuracy on ImageNet, surpassing the dense baseline by 1.8% while using 60% fewer FLOPs per token...",
  "table_captions": [
    "Table 1: Accuracy vs. compute comparison on ImageNet",
    "Table 2: Ablation study on sparsity patterns"
  ],
  "source": "cache"
}

{
  "owner": "example",
  "repo": "sparse-attention",
  "readme": "# Sparse Attention\n\nEfficient sparse attention implementation...",
  "file_tree": [
    "README.md",
    "train.py",
    "models/sparse_attn.py",
    "configs/imagenet.yaml"
  ],
  "tree_truncated": false
}

{
  "bibtex": "@article{2401_04088,\n  title={Attention Is All You Need (But Not All You Get)},\n  author={A. Researcher and B. Scientist},\n  year={2024},\n  journal={arXiv preprint arXiv:2401.04088},\n  url={https://arxiv.org/abs/2401.04088}\n}",
  "count": 1,
  "not_found": []
}

{
  "papers": [
    {
      "arxiv_id": "2503.12345",
      "title": "State Space Models Strike Back",
      "rank_score": 0.91
    }
  ],
  "total": 20
}

{
  "query": "diffusion models",
  "search_type": "semantic",
  "authors": [
    {
      "id": 12345,
      "name": "Jascha Sohl-Dickstein",
      "h_index": 28,
      "total_papers": 42,
      "primary_field": "cs.LG",
      "recent_paper_count": 7,
      "author_rank_score": 0.89,
      "research_topics": ["diffusion", "generative models", "score matching"],
      "semantic_scholar_id": "1234567",
      "years_active": 12,
      "relevance_score": 0.91
    }
  ],
  "total": 1
}

{
  "queried_author_ids": [12345, 67890],
  "window_years": 10,
  "edges": [
    {
      "from": 12345,
      "to": 67890,
      "papers_count": 5,
      "last_collab_year": 2024
    },
    {
      "from": 12345,
      "to": 11111,
      "papers_count": 2,
      "last_collab_year": 2023
    }
  ],
  "edge_count": 2
}

{
  "id": 12345,
  "name": "Jascha Sohl-Dickstein",
  "h_index": 28,
  "total_papers": 42,
  "total_citations": 18400,
  "avg_novelty_score": 0.71,
  "peak_novelty_score": 0.95,
  "author_rank_score": 0.89,
  "author_alltime_score": 0.84,
  "primary_field": "cs.LG",
  "recent_paper_count": 7,
  "venue_score": 0.82,
  "code_score": 0.65,
  "years_active": 12,
  "semantic_scholar_id": "1234567",
  "research_topics": ["diffusion", "generative models", "score matching"],
  "rank": 142,
  "top_papers": [
    {
      "arxiv_id": "2006.11239",
      "title": "Denoising Diffusion Probabilistic Models",
      "llm_novelty_score": 0.95,
      "year": 2020,
      "citation_count": 9800,
      "rank_score": 0.97
    }
  ]
}

{
  "author_id": 12345,
  "author_name": "Jascha Sohl-Dickstein",
  "papers": [
    {
      "arxiv_id": "2006.11239",
      "title": "Denoising Diffusion Probabilistic Models",
      "year": 2020,
      "citation_count": 9800,
      "rank_score": 0.97
    }
  ],
  "total": 42,
  "page": 1,
  "limit": 20,
  "has_more": true
}

{
  "query": "diffusion models for image generation",
  "paper_count": 50,
  "avg_similarity": 0.7823,
  "authors": [
    { "name": "Jascha Sohl-Dickstein", "paper_count": 4 }
  ],
  "publication_velocity": [
    { "month": "2022-01", "count": 3 },
    { "month": "2022-02", "count": 5 }
  ],
  "novelty_distribution": {
    "0.0-0.2": 2, "0.2-0.4": 8, "0.4-0.6": 18, "0.6-0.8": 16, "0.8-1.0": 6
  },
  "avg_novelty": 0.5842
}

{
  "embedding": [0.0124, -0.0381, 0.0562, "... 768 dims total ..."],
  "model": "gemini-embedding-001",
  "dimensionality": 768,
  "task_type": "RETRIEVAL_DOCUMENT"
}

{
  "topic": "diffusion models for protein structure",
  "papers": [
    {
      "arxiv_id": "2210.01776",
      "title": "Diffusion probabilistic modeling of protein backbone structure",
      "authors": ["B. Trippe", "J. Yim"],
      "published_date": "2022-10-04",
      "citation_count": 287,
      "rank_score": 0.84,
      "llm_summary": "Adapts DDPM to generate protein backbone coordinates...",
      "llm_novelty_score": 0.78,
      "similarity": 0.8312
    }
  ],
  "total": 15,
  "note": "Cheap retrieval only — no LLM synthesis. The /field-guide skill composes this with in-context reasoning."
}

{
  "id": "b7e3c1a4-2f5d-4c0a-9e1b-8a7c3d2e1f00",
  "status": "queued"
}

{
  "id": "b7e3c1a4-2f5d-4c0a-9e1b-8a7c3d2e1f00",
  "status": "queued",
  "request_counts": { "total": 3, "completed": 0, "failed": 0 },
  "results": null,
  "created_at": "2026-05-27T12:00:00Z",
  "started_at": null,
  "finished_at": null,
  "expires_at": "2026-06-26T12:00:00Z"
}